The effect of environmental stimulation on sensation seeking behavior of criminals and non criminals

Material Information

The effect of environmental stimulation on sensation seeking behavior of criminals and non criminals
Sensation seeking behavior of criminals and non-criminals
Brodsky, Annette M
Place of Publication:
Gainesville FL
Publication Date:
Copyright Date:
Physical Description:
vi, 88 leaves. : ill. ; 28 cm.


Subjects / Keywords:
Analysis of variance ( jstor )
Anxiety ( jstor )
Criminals ( jstor )
Mazes ( jstor )
Mental stimulation ( jstor )
Minnesota Multiphasic Personality Inventory ( jstor )
Physiological stimulation ( jstor )
Prisoners ( jstor )
Sensation seeking ( jstor )
Visual perception ( jstor )
Criminal psychology ( lcsh )
Human beings -- Effect of environment on ( lcsh )
bibliography ( marcgt )
theses ( marcgt )
non-fiction ( marcgt )


Thesis -- University of Florida.
Bibliography: leaves 71-75.
General Note:
Manuscript copy.
General Note:

Record Information

Source Institution:
University of Florida
Holding Location:
University of Florida
Rights Management:
Copyright [name of dissertation author]. Permission granted to the University of Florida to digitize, archive and distribute this item for non-profit research and educational purposes. Any reuse of this item in excess of fair use or other copyright exemptions requires permission of the copyright holder.
Resource Identifier:
000543418 ( AlephBibNum )
ACW7126 ( NOTIS )
13153457 ( OCLC )


This item has the following downloads:

Full Text








The author wishes to express her deep thanks to the many persons

who so generously contributed their time, money, advice, and encourage-

ment to this project.

Thanks are due to the advisory committee, Chairman Audrey S.

Schumacher, Louis Cohen, Marvin Shaw, William Wolking, and Bruce

Thomason for their contributions over the last seven years.

Special thanks are extended to the Southern Illinois staff members

who served as consultants to the project, Jacob Evanson, Jerry Hostetler,

and Alfred Lit. Appreciation for assistance is due to the experimenter,

Robert Nelson, and to John Bailey for the use of his films. I also want

to thank Clayton Ladd for the use of the Counseling Center facilities

and for time to work on the project, and Steven Danish for his moral


I am grateful to the Federal Bureau of Prisons for permission to

use the subjects and facilities at the Marion Federal Penitentiary, and

to William Lyle and Assistant Warden George Camp, for their help. At

the Illinois State Penitentiary, Menard Branch, I wish to thank John

Twomuey and Roy Kurtak. At all three institutions I am grateful to the

subjects who were kind enough to volunteer their time and efforts.

This study was partially funded by a grant from Research and Projects,

Southern Illinois University, and I am grateful to them for the support.

Above all, I am deeply indebted to my husband, Stanley Brodsky,

without whose encouragement and support this study never would have

reached completion.



ACKNOWLEDGEMENTS . . . . . . . . . . . ii

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


I. INTRODUCTION . . . . . . . . . . 1

Sensation Seeking as a Drive
Measuring Sensation Seeking
Survey of the Personality Correlates of Criminality
Sensation Seeking and the Antisocial Personality
Statement of the Problem

II. METHOD . . . . . . . . . . . 22

Measures of Sensation Seeking
Zuckerman Sensation Seeking Scale
Howard Maze Test
Perceptual Line Task
Measures of Personality Characteristics
Facilities Used
Sensory Deprivation Condition
Sensory Enhancement Condition
Order of Presentation and Instructions

III. RESULTS . . . . . . . . . . . 32

IV. DISCUSSION. . . . . . . . . . .. 54

Criminals as a Subject Population
Antisocial Behavior and Sensation Seeking
Adequacy of Sensation Seeking Tasks
Sensation Seeking as a General Factor
Limitations of the Measurement of Stimulation Conditions
Relationship of Anxiety to Sensation Seeking

V. SUMMARY . . . . . . . . . . . 68



BIBLIOGRAPHY . . . . . . . . 71

APPENDICES . . . . . ... . . . . . . 76

BIOGRAPHICAL SKETCH ...................... 88


Table Page

1. Means and Standard Deviations of Sensation Seeking Tasks 33

2. Correlations among Sensation Seeking Tasks . . . 35

3. Analyses of Variance of Mazes by Groups. . 0 . . . 38

4. Analyses of Variance of Sensation Seeking Scale by Groups. 38

5. Analyses of Variance of Lines by Groups ... 0 0 . .0 39

6. Analyses of Variance of Trends by Groups . . . 39

7. Analyses of Variance of Mazes by MMPI Pd Scale . . .. 41

8o Analysis of Variance of Zuckerman Sensation Seeking
Scale by MMPI Pd Scale . . a . o . . . . 41

9. Analyses of Variance of Lines by MMPI Pd Scale . . 42

10. Analyses of Variance of Trends by MMPI Pd Scale . . 42

11. Analyses of Variance of Mazes by MMPI Ma Scale . . .. 43

12. Analysis of Variance of Zuckerman Sensation Seeking
Scale by MNPI Ma Scale . . . . . . 43

13. Analyses of Variance of Lines by MMPI Ma Scale . . 44

14. Analyses of Variance of Trends by MMPI Ma Scale . . 44

15. Correlation of Sensation Seeking Tasks by MMPI Ma and
Pd Scores . . . . . . . . . . 46

16. T Score Means and Standard Deviations for MMPI Scales . 47

17. MMPI T Score Means for Marion Subgroups . . . . 49

18. The Relationship of High Points on Ma and Pd Scales
of the M1IPI . . . . . . . . . . 50

19. MMPI Pd High Points by Institution . . . 50

20. MIMPI Ma High Points by Institution . . 50

LIST OF TABLES (continued)

Table Page

21. Distribution of Sensation Seekers on the Zuckerman
Sensation Seeking Scale . . . . . . 51

22. Discriminating Items on Sensation Seeking Scale . 51



The objective of this study was to investigate the role of sensa-

tion seeking in the antisocial personality. Criminals were chosen as a

representative population of severe antisocial behavior. For the pur-

pose of this study criminal is defined as an individual who has been

convicted of a felony.

Sensation Seeking as a Drive

Hebb (1949) postulated that organisms seek an optimal level of

stimulation. Hebb's theory dealt with physiological :nachanisms, parti-

ularly neural circuits in the brain, that eventually formed into com-

plex patterns through learning and experience. These patterns result in

individual preferences by organisms for a certain degree of external and

internal stimulation, anything below or above which leads to an unpleas-

ant state of being that motivates the organism to reduce or increase the


Fiske and Maddi (1961) extend Hebb's notion of optimal stimulation

seeking to the human level. They hold that individuals operate accord-

ing to their own norms of internal activation. That is, everyone has a

level of physiological arousal or activation that is most effective for

him in dealing with his environment. This level of activation changes

according to the natural sleep cycle and according to the necessary

tasks at hand. Berlyne (1966) describes this fluctuating arousal level

as "a psychophysiological dimension, indicative of how 'wide awake',

'alert', or 'excited' an organism is at a particular time." To Berlyne,

arousal is closely related to drive and any factor that raises or lowers

arousal can also be seen as a factor that can increase and/or reduce

drive, thus enabling it to motivate behavior and give rise to changes of

behavior through learning.

Berlyne (1966) demonstrated some of these features in a study in

which he reports that rats subjected to noise between experimental

sessions had a higher arousal level than rats maintained in a quiet

room. Also relevant is a study by Glanzer (1953) in which rats were

confined to one arm of a T maze and then given a choice between the arms.

When confined for 15 to 30 minutes, they tended to choose the unfamiliar

arm more often than when confined for 1 minute.

Suedfeld (1967) has established evidence that stimulation seeking

does function as a drive in humans, having the same properties of reward

as does monetary incentive, under drive conditions of sensory restriction.

Other studies in sensory deprivation (Zubek, 1969a) have indicated the

incentive value of sensory stimulation during a period of deprivation.

These studies imply that arousal level can be altered through en-

vironmental manipulation which could have an effect on an individual's

behavior. That the optimal level is neither at the upper or lower ex-

treme of activation is suggested by Berlyne and Lewis (1963). They found

that the reward value of stimulation resulting from diverse exploration

is an inverted U-shaped function of the degree to which the stimulus

raises arousal. In another study, Berlyne found that the extent to

which a stimulus raises arousal increases with its novelty and with the

subject's initial arousal level. Animals given amphetamines performed

more responses with familiar reinforcing stimuli, whereas control

animals injected with a saline solution performed more responses with

a novel reinforcing stimuli (Berlyne, 1966). Thus, it would seem that

different levels of wakefulness, or arousal, are involved in the indivi-

dual's momentary need for stimulation as well as his more general, per-

haps constitutional variation from the norm.

Fiske and Maddi point out that the individual has some control over

his arousal level and can manipulate the amount of impact of stimulation

by many means. For Fiske and Maddi, impact is a function of the inter-

action of the intensity, novelty, and meaningfulness with which stimulus

input affects the individual. While some of these means are facilita-

tive to effective performance, others can be detrimental.

For example, the surgeon requires a certain narrow range of activa-

tion level within which he can function during an operation most effec-

tively. He must not be either too tense and excited, or too drowsy. He

can achieve this by lying down if he is tense, or by slapping his face

or splashing cold water on it if he feels drowsy. These methods are

facilitative to raising or lowering arousal level. There are times how-

ever when individuals are not aware of their need for more stimulation

and unconsciously behave in a manner to increase stimulation. It is at

times like this that detrimental means may come into play. For example,

during a monotonous task, there is not enough stimulation input for the

level of arousal needed to remain alert enough to complete the task,

particularly in a task that involves observation at all times in order

to catch the momentary stimuli that must be dealt with, i.e., a slow

assembly line. Given such a situation, subjects will either indulge in

feet tapping, humming, or other helpful measures to increase stimulation.

Others will daydream or focus on extraneous stimuli (looking out a win-

dow) that will distract them from the task and prove detrimental to per-


These last two examples have dealt with specific tasks and a parti-

cular motivation requiring a change in activation level. When there is

no specific goal, however, the organism, according to Fiske and Maddi,

tries to strive for a moderate level of activation that is most appropri-

ate for his present level of cyclical arousal. When the ongoing stimu-

lation is not especially intense or meaningful, then variation of stim-

uli plays an important part in satisfying the need for stimulation. For

example, while lying on the beach, many individuals find themselves

watching the ocean waves rise and fall, or when sitting quietly at home

people will spend long periods of time satisfied to watch the variation

of a fire in the fireplace.

Learning depends on a certain level of activation that may well

differ for individuals. According to Fiske and Maddi, there is an in-

verted U-shaped function between activation level and effectiveness of

performance. Stennett (1957) has shown that under moderate levels of

skin conductance and muscle tension subjects learned better on an audi-

tory tracking task than under different instructions that yielded higher

or lower levels of physiological arousal. Poor performance is associa-

ted with low activation levels due to a lack of sufficient cortical

arousal to permit relevant responses to continue at a normal rate.

Efficiency is reduced when the organism is unable to raise its level of

activation toward the optimal zone for the task at hand. A similar

phenomenon occurs at the other extreme. When there is a very high level

of activation the organism may find itself disrupted from his ability

to concentrate on the particular task because of the bombardment of

excess stimuli. White (1959) states that active stimulation seeking

is an ever present need but it functions only in the absence of other

motivational needs. Specific motives become prepotent and serve to

direct the activity of the organism from his otherwise state of free

exploration and interaction with the environment. Thus, we might

suspect that with different physiological norms or levels of arousal,

different individuals would find themselves more or less preoccupied

with motivational needs while others would be free to seek additional

stimulation and in some cases, in a great need for stimulation that is

hard to fulfill in our present social order. It is perhaps easier for

us to see the difficulties confronted by those at the low end of the

stimulation seeking continuum, such as the tense individual who is con-

stantly driven by unsatisfied primary needs which restrict, his oppor-

tunity for stimulation seeking behavior. Berlyne (1966) has shown that

people are less eager to seek stimulation under very high levels of

arousal, such as hunger or pain.

However, the person who operates at a low level of anxiety, as pre-

sumably does the psychopath, may also be operating at a low basic arousal

level and feel a frequent need for additional stimulation. For him, the

physiological concomitants of anxiety might add to the effectiveness of

his performance on motivational tasks and increase his learning effici-

ency as did the epinephrine injections in the psychopaths in the study

by Schacter and Latane (1964). Their psychopaths learned poorly under

avoidance conditioning when the motivating conditions were the same as

normals, but did remarkably better under the influence of the sympathe-

tic nervous system stimulant, epinephrine. Leuba (1955) has commented

on the need to concern ourselves with the learning of individuals who

require greater than average stimulation as reinforcement. "Quite

different procedures are indicated if learning in daily life occurs

rarely, rather than ordinarily, through tension reduction. The teacher

may have to increase tensions and make the school situation a more ex-

citing one."

Measuring Sensation Seeking

Several instruments proposing to measure sensation seeking or rela-

ted phenomena exist in the literature. They can be divided into three

main categories questionnaires, performance tasks, and perceptual


Questionnaires which are dependent on the verbal awareness as well

as honest disclosure of the subject are predominant. Penney and Reinehr

(1966) report a Stimulus Variation Seeking Scale for Adults which deals

with external stimuli and is mainly cognitive rather than sensory. High

scorers on this test are also prone to perceive increased movement on

the autokinetic effect. The test correlates with Zuckerman's Sensation-

Seeking Scale .33 (p<.05). Also, Garlington and Shimota (1964) report

a Change Seeker Index, based on the need for variation in one's stimulus

input. This instrument is closer to behavioral ratings used by Luebeck

(1967) to identify sensation seekers by their activity in confinement.

Other related questionnaires can be found in the sensory deprivation

research to identify symptoms and reactions to isolation (Zuckerman,


Perhaps the most promising questionnaire measure of sensation seek-

ing is Zuckerman's Sensation Seeking Scale (SSS). Zuckerman (1964) de-

veloped his instrument to measure optimal sensation seeking. His pur-

pose was to include the various aspects of sensation seeking and to de-

rive a general factor. He included items dealing with (a) preference

for extremes of sensation, such as cold, heat, noise, etCo, (b) prefer-

ence for the new versus the familiar, (c) for irregularity versus rou-

tine, (d) enjoyment of danger, thrills, (e) social unpredictability,

(f) adventure versus security, and (g) a need for general excitement.

Some of the correlates of this scale are: positive with education, in-

telligence and perceptual ability, and negative with age (Kish and

Busse, 1968). However, the correlations only hold at the below average

levels indicating that a certain minimum level is necessary for the task

to discriminate between individuals on sensation seeking. In terms of

age, the effect is noted above 35 years. Farley and Farley (1967) re-

port a .47 correlation with the Extraversion scale on Eysenck's Person-

ality Inventory, and Zuckerman (1968) reports a positive relationship

with F, Pd, and Ma on the Minnesota Multiphasic Personality Inventory

(MMPI)o In a recent unpublished study (Althoff, personal comiLunication)

a .60 correlation between the SSS and a scale reporting interest in,

knowledge and use of psychedelic drugs was found. Also, Brownfield

(1966) found that mental patients in general were low sensation seekers

as compared to normals. Under deprivation, the high scorers showed more

discomfort, anxiety, boredom, and disorganization (after 24 hours),

while the sensation avoiders reported pleasant experiences and occasion-

ally asked to remain longer.

Of the performance tasks, mazes have proved most fruitful, whether

on the subhuman level (Berlyne, 1966) in measuring exploratory drive on

the human level as reported by Howard (1961). Howard developed his

mazes for the purpose of measuring change seeking behavior in a simple

task that would be useful in research on sensation seeking. Other

mazes had previously been used to differentiate between personality

groups, notably Porteus Mazes which have been found to be the most

consistent performance tests in differentiating between delinquents and

nondelinquents (Waldo and Dinitz, 1967). Docter and Winder (1954),

Fooks and Thomas (1957), and Gibbens (1958) all found differences be-

tween delinquents and controls on the Porteus R scores. q is a quali-

tative scoring procedure that evaluates expressive movements of the

subject. Porteus states that q is a measure of haphazardness, impulsiv-

ity, and failure to pay attention to other considerations in the subject's

haste to get done. The various authors have interpreted it to be a

measure of impulsivity since three of the four discriminating components

of the total of nine are based on the subject's disobeying instructions

of some type. Gibbens (1958) found a complex relationship between

delinquency, Porteus q, and the MMPI. q correlated with Hypomania (Ma),

which did not differentiate his two populations, but q did not correlate

with Psychopathic deviate (Pd), which did differentiate the two groups.

Howard (1961) developed a maze test with the purpose of providing

a simple task to be used by researchers in the measurement of sensation

seeking. His test did not correlate with Zuckerman's scale (Zuckerman,

1964), but it has shown some construct validity in relation to certain

factor traits on the 16 PF Questionnaire. These factors and the

direction of correlation are (a) negative with 0, a factor testing timid-

ity and proneness to guilt, (b) negative with Q4, a factor of high con-

trols and low drive energy and (c) positive with kQ1, a factor indicating

a tendency to experiment with problem solutions, resourcefulness, and


On a perceptual level of measuring sensation seeking need, adapta-

tion phenomena are predominant. The autokinetic effect as a need for

stimulation is defined by Berggren (1968) and an autokinetic tracking

task was used by Rein.ehr to validate his own sensation seeking ques-

tionnaire (1965). Luebeck (1967) used auditory tone decay to test

sensory adaptation, but without promising results.

Petrie (1967) has used a kinesthetic task to differentiate the

perceptual reactance of individuals into cognitive styles. She used

a size comparison task that has successfully differentiated between

various populations. Those who exaggerate the size of the comparison

stimuli she refers to as augmenters. Those who tend to report the size

as smaller are called reducers, and the middle group, moderates.

Petrie suggests that the kinesthetic effect of reduction and aug-

mentation will also apply to other modalities. In a pilot study she

demonstrated the comparison of the kinesthetic task with one of compar-

ing weight of objects and found a significant relationship. Also, she

states that there is some evidence for light intensities to be augmented.

When auditory stimulation and thermal stimulation are presented before

administering the task, there is increased kinesthetic reduction.

The main findings over many populations is that augmenters are

unusually sensitive to pain and other strong stimuli, while reducers

are exceptionally tolerant of pain. At the other end of the stimulus

continuum, Petrie noted that the augmenters were not particularly dis-

turbed under sensory deprivation situations and tended to be those who

would volunteer and remain under restrictive conditions for long periods

of time. The reducers, however, were those who quit first, made remarks

that the situation was very stressful, and stated that they would prefer

physical punishment in preference to returning to the deprivation situ-

ation (in this case being placed in an iron lung). Petrie also notes

that reducers predominated in delinquent groups, augmenters in neurotic

and psychotic groups.

The perceptual style did not seem to be a matter of different

thresholds of stimuli but rather the reaction of adaptation to the stim-

uli after it occurred. Pain threshold and tolerance for pain are not

the same. The augmenters although they had low tolerance did not

necessarily have low thresholds. Also, the reducers' tolerance of pain

seemed to be independent of the cause of the pain. Petrie believes that

the processing of pain is determined by the perceptual reactance of the

person which determines whether the sensation will be diminished as in

the reducer or magnified as in the case of the augmenter. However, other

investigators (Zubek, 1969a) have found thresholds for pain sensitivity

to be very readily changed under stimulus deprivation conditions. Sensi-

tivity to pain after only 5 minutes of deprivation has been reported to

increase and then show adaptation in an oscillatory pattern of decreased

and then increased sensitivity. Another objection to Petrie's inter-

pretation of her results is that performance of the task by a blindfold-

ed subject is in effect testing the subject under restriction of visual

stimulation and thereby introducing a sensory deprivation experience

that may be affecting the perceptual need state.

Survey of the Personality Correlates of Criminality

Criminal activity is often associated with the concept of a parti-

cular "criminal mind." The offender against society's rule frequently

is seen as a person with a defect of personality that permits him to

indulge in such otherwise unacceptable behavior. Historically, severe

antisocial behavior has been labeled as "moral insanity'" and jurisdic-

tional systems have made provisional tests to judge these persons as to

their mental capacity to adhere to normal, moral behavior. Legal dis-

tinctions, however, can only respond to severe moral deficits. The rec-

ognition that lesser degrees of difficulty with impulse control could

present problems to society and to the individual has been investigated

through the avenue of psychological research into personality variables.

One diagnostic category frequently used in research on severe antisocial

individuals has been the psychopath.

Until recently, the psychopath has presented difficulties to inves-

tigators because of the wide variety of symptoms that were often included

in the total picture. This diagnostic label has come to stand for some

fairly clearly defined patterns of behavior, however, due largely to the

writings of Cleckley (1959) and the McCords (1964). In their review of

the literature, the McCords state that probably all social scientists

would agree on the following characteristics of psychopathy. "The psy-

chopath is asocial, shows aggressive behavior, extreme impulsivity, little

or no anxiety or guilt and an inability to form lasting bonds of affec-

tion with others." Cleckley (1959) adds, "he fails to grasp the meaning

of responsibility, makes repeated and inadequately motivated aggressive

acts, does not profit from experience, and lacks insight into his own


Although there may be general agreement on the major characteris-

tics of the psychopath, occasionally the elimination of one or another

symptom can make a significant difference in the expected actions of the

individual. For instance, some investigators consider lack of anxiety

of paramount importance (Schacter & Latane, 1964; Lykken, 1957). Others

attend more to the actual behavior of impulsive acts, ignoring the auto-

nomic activity displayed, or the emotional demeanor demonstrated. This

is most evident in studies using prison inmates or delinquent popula-

tions (Waldo and Dinitz, 1967).

Many investigators have identified psychopaths on the basis of test

scores, particularly MVPI patterns with significantly high Psychopathic

Deviate Scales (Pd) and/or Hypomania (Ma) Scales. This is perhaps the

most fruitful and reliable measure for identifying the group (Waldo &

Dinitz, 1967). In an earlier review of the literature on personality

traits of criminals, Schuessler and Cressey (1950) concluded that "the

evidence favors the view that personality traits are distributed in the

criminal population in about the same way as in the general population."

However, since 1950, 81% of the studies surveyed by Waldo and Dinitz re-

ported significant differences. The MMPI was used 29 times in this sur-

vey (31% of all studies) and in 28 of the 29 studies, Pd scores consis-

tently discriminated criminals from noncriminals. The Pd scale was

developed through the differentiation of prisoners and delinquents vs.

normal and therefore represents a more global measure of antisocial

behavior than the strict syndrome of the "primary psychopath." The

items of the Pd scale include experiences dealing with infractions of

the law, social maladjustment and absence of pleasant experiences. The

Ma scale represents the acting out potential of this group with items of

expansiveness, egotism and irritability. Thus in these studies the two

scales together indicate a group of antisocial, acting out individuals

who were not successful in staying out of prison.

The "psychopath" it seems does not necessarily land in prison, and

the criminal (or prisoner) is not always a "primary psychopath." It is

not uncommon for college students to exhibit peaks on Pd and Ma, but

this is not frequent or at a high degree of elevation (Lanyon, 1968; Fry,

1952). The habitual criminal is more likely to be an individual who,for

one reason or another, has difficulty with impulse control and readily

indulges in antisocial acts that transgress the law. In the case of the

recidivist, this behavior takes on a more meaningful urgency, in that he

has demonstrated an inability to "learn by experience" that crime does

not pay.

This inability of the habitual criminal to change his pattern of

behavior or to become motivated to lead a more normal life in spite of

his verbalizations to the contrary has suggested to researchers that a

learning deficit exists. Psychopaths, however, have been shown to learn

basically as well as other groups under most conditions. Gurvitz (1947)

and Kingsley (1960) found that they perform normally on intelligence

tests. Fairweather found that they were not inferior on serial rote

learning tasks (1953) and Sherman (1957) discovered no inferiority in

their retention of learned material. However, psychopaths have consis-

tently shown deficits in their ability to become conditioned, particu-

larly where anxiety, fear, or secondary reinforcement is involved.

Johns and Quay (1962) found that psychopaths were significantly poorer

than other groups on verbal conditioning and appeared to be unmotivated

by social reward. Lykken (1957) found that the "psychopath" has diffi-

culty acquiring classically conditioned fear responses.

On the other hand, Persons and Bruning (1966) found that their

psychopaths could learn an instrumental task as well as others under

the proper conditions of motivation. Their subjects had volunteered for

the experiment in lieu of their otherwise boring prison routine. The in-

crease of stimulation apparently served as the motivator. Weisen (1965)

found that when increase of stimulation was the reward, psychopaths

could be instrumentally conditioned, while the anxiety neurotics in his

experiment were conditioned by the decrease (offset of the stimuli).

Schacter and Latane (1964), who replicated Lykken's work on demonstrating

a lack of anxiety in psychopaths, extended the study and found that the

psychopaths' learning improved greatly under the influence of epinephrine.

So, from learning studies using psychopaths, we see that the psychopath

is deficient only when the motivators involve anxiety or when the reward

involves social reinforcement. When excess stimulation is the reward,

or when arousal is artificially increased during the task by means of

drugs, the psychopath actually performs better than normals. The rela-

tionship between anxiety, social reinforcement and stimulation seeking

is unclear and individually these factors need more exploration. The

present study, however, will deal only with the third factor, the rela-

tionship of sensation seeking to criminal behavior.

Sensation Seeking and the Antisocial Personality

In her study of delinquents, Petrie (1967) found that there were

significantly more reducers than augmenters. Thus, the delinquents

followed the reducers' pattern of being tolerant of pain (all delin-

quents who had tatoos were reducers), particularly uncomfortable in

sensory deprivation situations, and susceptible to strong satiation

(adaptation) when the perceptual intensity of a strong stimulus over

a period of time becomes diminished by the individual. For the redu-

cer, then, restriction of stimuli often represents a stress or suffering.

Fiske (1961) reports that in sensory deprivation, stress is apparent by

a rise in the secretion of epinephrine and norepinephrine. Petrie sug-

gests the feeling of stress that the reducer must feel: "In order to

hang onto his feelings of reality he becomes greedy for increased amount

of sensory input of any kind." Thus, he is vulnerable to strong sensa-

tions of both positive and negative appeal, preferring his rewards and

punishments to be of an active nature.

Weisen (1965) found that psychopaths (subjects high on MMPI Pd and

Ma scales) were significantly more reinforced by onset of stimulation

than anxiety neurotics (subjects high on MMPI D and Pt scales and low on

Pd and Ma) who were reinforced by the offset of stimulation. These sub-

jects were instrumentally conditioned in a Taffel-type verbal condition-

ing situation. The stimulation variable used was music piped into the

room when a lever was pressed.

Quay (1967) has proposed a two-factor theory to explain the basis

of what he considers a pathological degree of sensation seeking in

psychopaths. One factor reflects the adaptation phenomenon that Petrie's

work implies. Quay cites other studies that can also be interpreted in

this vein, most dealing with GSR studies and other physiological measures.

The second factor involved in Quay's proposal is the initial reactivity

to new stimuli. He proposes that the psychopath is in a deprivation

state because he is subjected to lowered arousal reactivity and thereby

misses the needed stimulation to function adequately. Since psychopaths

also tend to adapt quickly, they find themselves back in a deprivation

state sooner than normals. Lowered arousal reactivity could also ex-

plain the lack of anxiety of the antisocial personality as the physio-

logical concomitants of anxiety do not have the opportunity to develop

within the time span of the individual's period of reaction. Research

related to the issue of reactivity, however, is inconsistent. It

appears upon close inspection of these studies that one major reason

may be the populations involved. When "psychopaths" are so-called be-

cause of lack of anxiety and remorse there are often different results

than when the main consideration is the behavior itself being of a high-

ly impulsive nature. These studies dealing with impulsive "psychopaths"

seem to be in the direction of the high reactivity results. However,

in relating reactivity to sensory deprivation research, Zubek (1969b)

notes that "quitters" tend to have an abnormally low baseline level of

adrenalin before isolation and when tested months later they also show

a rise in noradrenalin. This occurs only in those who leave the exper-

iment early due to stress and in those who have experienced unusually-

severe sensory deprivation conditions such as water immersion.

Leubeck (1967) attempted a test of Quay's theory. His psychopaths

did not adapt quicker to an auditory tone decay situation and they did

not show significantly different reactivity from normals. Although he

had some clinical evidence from raters of the psychopathic group's be-

havior that they were more active and prone to do things that could be

labeled as "sensation seeking',' they did not reveal this on the Zucker-

man scale. Leubeck had some methodological problems, however, that could

have accounted for this lack of results. For- one, he equated threshold

with reactivity and tested for auditory thresholds. Petrie, however,

has noted that the reducers in her experiments were not particularly

different in their pain thresholds although their tolerance for pain

differed from augmenters. In fact, there is evidence that if a differ-

ence in perceptual thresholds exists between psychopaths and normals,

the psychopaths appear to have lower thresholds. Glaudin (1954) found

that on a word recognition task presented with a tachistoscope psycho-

paths had significantly fewer exposures necessary to identify words. He

concluded that psychopaths were perceptually vigilant and we might ex-

tend the interpretation that the need for stimulation encourages a vigi-

lant attitude in the perceptual style.

Skrzypek (1969) studied the effect of perceptual isolation and arousal

on the report of anxiety and preference for complexity and novelty in

neurotic versus psychopathic delinquents. Selected on the basis of a

behavioral rating checklist developed by Peterson and Quay, he found a

significant negative correlation between novelty preference and anxiety

level, and between complexity preference and anxiety. The psychopathic

delinquents were higher on both novelty and complexity. After isolation

for40 minutes, psychopathic delinquents increased their complexity scores

significantly more. After arousal, the neurotics increased their anxiety

scores and decreased on complexity.

This study supports the negative relationship of anxiety and psycho-

pathy and shows the effect of deprivation of stimulation on increased

preference for stimulus variation (novelty and exploration). On behav-

ioral levels more basic than variability will this apparent need for

stimulation operate on a motor performance task (Howard's mazes) or on

the verbal report level, or the perceptual level? The short time differ-

ences indicate the lack of necessity for prolonged deprivation, which

often adds other variables such as anticipation, sleep, soreness, and

illegal activity (Zuckerman, 1969a).

Statement of the Problem

The association of sensation seeking and antisocial behavior has

not been established although there is strong evidence that the relation-

ship may be high. The predominant personality profile of psychopaths on

the MMPI with elevated Pd and Ma Scales has been found to be character-

istic of those who tolerate sensory deprivation poorly (Myers, 1969).

Delinquents have been shown to react with stress under sensory depriva-

tion (Petrie, 1967) and also to make positive scores on tasks which

measure preference for complexity and novelty (Skrzypek, 1969). Common

correlates of sensation-seeking and psychopathic personality traits are

also supportive of the conclusion that traits such as impulsivity, ex-

traversion, insolence, perceptual reactance leading to rapid adaptation,

etc. may be attributable to a need for sensory stimulation. Supporting

studies have been discussed previously in relationship to the sensation

seeking tasks for which they served as validation.

The purpose of this study was to explore directly the relationship

between the antisocial personality and the need for stimulation, through

a study of a criminal population. The convicted and incarcerated prison-

er in a maximum security prison has demonstrated behaviorally that in the

face of great risk, he has still found the motivation to perform a se-

verely antisocial act. This action is not necessarily performed by

"psychopaths" only, and therefore, it was felt important also to inde-

pendently study the personality profiles of all subjects.

The study investigated the relationship of sensation seeking and

criminality across three types of tasks purporting to distinguish indi-

vidual differences in sensation seeking. It was also concerned with

comparing prisoner and nonprisoner populations on their reactions to

these tasks as a function of their momentary level of stimulation. In

addition to theoretical interest, these relationships have meaning for

the advisability of confinement as punishment, and for restrictive

procedures during rehabilitation, that may well put extraordinary stress

on those who have a greater than average need for stimulation in daily


The question of the generalizability of sensation seeking from one

sensory mode to another, as well as its generability over levels of be-

havior has not been clear. It was predicted on the basis of previous

evidence in the individual tasks that there would be a general factor

of sensation seeking. The three tasks chosen as representative of

different levels of sensation seeking were Zuckerman's Sensation Seeking

Scale (a verbal report scale), Howard's Mazes (a non-verbal performance

test), and a modification of Petrie's perceptual adaptation task.

Howard's Mazes test measures a type of exploration consisting of

change behavior in a simple, boring task. It was predicted that crimin-

als, if they do indeed operate on a higher need for stimulation at all

levels of arousal, will demonstrate a higher score on this task.

Zuckerman's Sensation Seeking Scale claims to be diverse in its

areas of sensation seeking, and is a self-report of experience and be-

havior. An individual's awareness of and stated preferences for choices

on a questionnaire may be an important issue in his understanding of his

motivations. In view of Zuckerman's negative results in correlating his

scale with Howard's Maze Test, as well as Luebeck's difficulty in estab-

lishing any relationship between the SSS and ratings of prison behavior

in his subjects, prediction of high scores for criminals on his scale

is somewhat tenuously made.

Petrie's notion of categorizing individuals into perceptual types

according to their exaggeration or diminution of stimuli related to their

"stimulus hunger" is the basis for her task being seen as a measure of

stimulation seeking. It was predicted that a perceptual task of this

nature would serve as another possible behavioral level of sensation

seeking, with reducing being related to criminality.

The following hypotheses were tested:

I. Criminals will be higher on sensation seeking tasks than

will noncriminal controls.

A. They will attain higher scores on the Zuckerman Sensation

Seeking Scale.

B. They will make more changes in paths on the Howard Maze


C. They will show a constant error in the direction of

reduction of size of a drawn line or circle over five

continuous trials.

This hypothesis was based on the evidence of Weisen (1965), Skrzypek

(1969), and Petrie (1967) that psychopaths and delinquents behave in a

manner to increase stimulus input. Part C of the hypothesis was based

on an indirect measure of sensation seeking developed by Petrie (1967).

J.I. Sensory restriction conditions will increase the amount of sen-

sation seeking on the three experimental tasks. This hypothesis was

based on Fiske and Haddi's theory (1961) which states that the organism

strives to reach an optimal level of stimulation and will seek to modify

the stimulus input level to match his current level of arousal.

III. There will be an interaction effect, with deprivation having

a greater effect on criminals than on controls. This is based on the

notion that the same short period of deprivation represents more stress

to the criminal group and will result in a greater motivation to increase

stimulation by that group (Schacter, 1957; and Zubek, 1969).

IV. There will be no differences in direction of the results for

the various tasks. This is based on the conclusion of Petrie (1967) and

Zuckerman (1968) that sensation seeking is a general factor demonstrable

over various levels and across sensory modes.

V. The subjects with high points on Pd or Ma on the M'1PI will tend

in both criminal and control groups to be higher sensation seekers on

the three tasks than those who show other scale high points. This is

based on findings by Myers (1969) that the volunteer for sensory depri-

vation experiments (over 500 in his population) is lower on Hs, D, lHy

and Pt. This is also based on his survey that in 5 studies involving

MMPI and sensory deprivation, those who tolerate the experiments poorly

are high on Pd, Ma, and measures of aggression.




The experimental group came from two prison populations. Forty

subjects were inmates at the Illinois State Penitentiary, Menard

Branch. Nineteen subjects were incarcerated at the Marion, Illinois

Federal Penitentiary. Both prisons are maximum security institutions

in southern Illinois. The prisoners were selected for participation

if they met the criteria of being under 30 years of age and enrolled

in college level courses at the institution. An individual interview

with the author was held for each prisoner who met the criteria and

they were asked to volunteer for the study. They were told that it

was not dangerous, no shock was involved, no individual results would

be available to the institution, and that they might or might not during

the one hour experiment be taking some short, verbal tasks, paper and

pencil mazes, watching a movie, or reclining on a lounge chair. The

nature of the study was not disclosed until after all subjects had been

tested. The letter of explanation is reported in Appendix C. Two

subjects at the Marion Penitentiary and one subject at Menard elected

not to volunteer. A letter of appreciation for participation was

placed in the files of those who completed the experiment.

The control group consisted of 60 undergraduate students at

Southern Illinois University. All were enrolled in the introductory

psychology course, Spring term, 1969. They received credit for

participation in terms of added points to their final grade in the

course. Only male students under 30 years were accepted for the

study. On the sign-up sheet for volunteers the study was entitled,

"Mazes and other tasks'.' Any subject who asked for further explana-

tion was given information equal to that told the prisoners. An

explanatory letter (Appendix D) was mailed to all participants upon

completion of the data collection.

Measures of Sensation Seeking

The Zuckerman Sensation Seeking Scale: This scale is described by

Zuckerman (1964) and is a 34 item questionnaire with pairs of alter-

native choices making up each item. The subject is given a forced

choice and the score is the number of positive choices for the member

of each pair that refers to a sensation seeking activity. Zuckerman's

criterion for a sensation seeker is a score of 15 or higher.

The original 54 items were selected to pertain to preference for

extremes of sensation (heat, cold, noise, etc.). Others pertained to

novelty versus familiarity, irregularity versus routine, thrills, inter-

personal unpredictability, adventure versus security, and general excite-


The items with splits greater than .85-.15 for the total popula-

tion to which it was administered (268 male and 277 female undergraduate

college students) were then eliminated and the remaining items were

factor analyzed separately for each sex. One large factor emerged. A

revised 34 item scale was selected based on items with loadings of .3

or higher on this factor. Males and females were given only items which

were significantly loaded for their own sex. Split-half reliability

of male and female scales showed r's of .68 and .74, respectively.

The 8 items for females only were included and identified in the item

analysis (Appendix B).

Validation of the scale has been primarily by correlation with

other probable sensation seeking traits. To date the following have

shown to be significantly correlated with the revised 34 item scale:

persons who volunteer for unusual experiments, quitters in long dura-

tion sensory deprivation experiments (Zuckerman, 1966), Eysenck's

Extroversion Scale (Farley, 1967), MMPI, F, Pd, and Ma Scales, In-

solence Scale, asocial direction of Gough's Socialization Scale, and

a preference for complexity and tolerance for ambiguity (Zuckerman and Link,

1968). It is also negatively correlated with anxiety, and shows no

correlation with Howard's Maze Test (Zuckerman, 1964).

Howard Maze Test

This test consists of two alternate forms, one providing alternate

paths to the same goal and the other, alternate paths to alternate goals.

The two forms are equivalent. The intercorrelations vary from .42 to

.67 (Howard, 1961). The mazes were developed in order to measure change-

seeking behavior, or spontaneous alternation, by means of a paper and

pencil task. Originally, three forms were used but one was eliminated

as a result of poor equivocality with the other two. All paths of each

of the forms are equally correct and each maze has 10 choice points. A

packet of five identical mazes is presented on each trial. Change is

defined as the number of segments traversed on a particular presentation

that differ from the corresponding segments traversed on the maze

immediately preceding it. The Total Change Score reflects the amount

of change for each adjacent maze in the series, and the Stimulation

Seeking Score is a weighted, final score emphasizing change from maze

to maze within a series (Howard and Diesenhaus, 1.965). The lowest

score possible is 0, and the highest is 16 (10 segments on each maze x

four adjacent mazes).

Howard typically gives five mazes to each subject and to conceal

the purpose of the test he uses it as a filler between other tasks.

While he has been able to get reliable test-retest data, Zuckerman

(1964) has noted in use of the mazes that subjects often rush through

the task with the notion that the task to follow, yet to be completed,

would be more efficiently remembered. Whether or not the high motiva-

tion to do it as quickly as possible makes a difference is unclear, but

one could hypothesize that the anxious person would be more motivated

to hurry, and the psychopath, free of anxiety, would be more willing to

spend time exploring and thus achieve a higher score in the direction

of change seeking.

The construct validity for the Howard mazes is partially established

by Domino (1965) who found a positive relationship for college students

between the test and number of activities attended. Also, Sidle (1963)

found a negative relationship to social withdrawal in psychiatric patients,

and Howard and Diesenhaus (1965%) found the test to correlate negatively

with the 16 PF Personality Test dimensions of 0- timidity and guilt

proneness, Q4- anxious and high energy, and positively with 0l-experimenting

nature. Howard (1965b) also reports on Form A, internal consistency re-

liability (estimated by alpha) of .69 for their total age group (202

elementary, high school, and college students), and .76 for the 77

college students, alone. Domino (1965) found retest reliability after

one week of .76 correlation of 27 of his original subjects, and .71

correlation after six months.

Perceptual Line Task

Petrie (1967) developed a specific kinesthetic task procedure.

Two tests are given. They are separated by 48 hours and prece ded by

a 45 minute rest period. In the large block comparison, a 2 inch

block is the stimulation, a 1 inch block the measurement. In the small

block comparison, the 1 inch block is the stimulation, a 2 inch block

the measurement. The subject is blindfolded and asked to feel the

measurement block with one hand and estimate its width with the other

hand on a tapered block. Four trials are given. He is then asked to

rub the stimulation block for 90 seconds and then repeat the estimation

of the measurement block. Additional rubbing of 90 seconds and 120

seconds are then tested in the same manner.

Results indicating Augmenting show that the subject, by the end of

the rubbing period, perceives the measuring block as about 50 7%

increased in size, while the Reducer has decreased the size about

50 % Dividing her non-pathological group into thirds, Petrie chose

as the critical points of differentiation an average reduction of 5.4 mm

or more for Reducer, increase of 5.4 mm or more for Augmenters and those

falling between formed the Moderates. The Reducer, she notes, decreases

subjective size whether the stimulating block is smaller or larger than

the measuring block and the Augmenter likewise enlarges both.

Split-half reliability was tested by correlating the sum of the

first and fourth difference scores with the sum of the second and third

difference scores for each test session. Product moment correlation

was .979 for children subjects and .973 for student nurses.

A perceptual task to measure the size constancy and adaptation

effects that Petrie (1967) uses in her kinesthetic task was devised

by the author. The essential ingredients of such a task were that the

subject produce an estimate of perceptual size and that he make successive

judgments in order to measure adaptation. Two related tasks were used

as alternate forms. Both tasks were given to all subjects. The drawing

of a line estimated to be three inches long or the drawing of a circle

estimated at three inches in diameter were used as tasks. The subject

was blindfolded for these tasks and given five successive trials with

5 seconds between trials. The accuracy score of the subject was the

mean length of the five lines or mean diameter of the circle in inches.

A trend score for length of line was computed by weighting the individual

means from the first to fifth trial and this was considered a measure of

reducing or augmenting as an adaptation phenomenon.

Measures of Personality Characteristics

The Minnesota Multiphasic Personality Inventory was used as the

traditional personality measurement against which the groups could be

compared independently from their institutional classification as an

aid in understanding the result. This instrument is a 500-item, True-

False questionnaire, with ten clinical scales and three validity scales.

The only instructions given were those printed on the booklet unless

further questions were asked. The reliability and validity of this

test is widely established and it has been used on prisoner populations

as well as college students and other clinical groups (Dahlstrom and

Welsh, 1960; Waldo and Dinitz, 1967; and Lanyon, 1968).


Facilities Used

All groups were tested at their own institution. In each case,

three small adjacent rooms were used for the pre- and post-testing and

the sensory restriction condition. A larger room was used for the film

presentation. The same male experimenter was used for all subjects, a

male college senior, who was trained in the administration of the ex-

periment by the author.

Sensory Deprivation Condition

Subjects were tested in a small, darkened room on an aluminum

chaise lounge with a foam pad, in a reclining position. Blindfolds

were made from gauze strips of four thicknesses with dark silk strips

inside to further block out light. Ear plugs covered by headphones

over the ears were used to deaden sound. The hands were loosely

wrapped in cloth to decrease motor activity such as drumming fingers.

Subjects were asked to remain still and quiet for 30 minutes.

Sensory Enhancement Condition

Laurel and Hardy comedies were shown for 30 minutes at a loud

volume. The particular film shown varied according to their avail-

ability, but all were chosen to be 30 minute films with sound tracks

that did not have any particular content associated with criminal be-

havior. The films used were: "Any Old Port' "Some Fine Mess" and

"Chickens Come Home" The films were shown to groups of three and

subjects were told that they could talk if they wished, as long as

they did not discuss the experiment.

Order of Presentation and Instructions

All groups were randomly split into either experimental (sensory

restriction) or control (sensory enhancement) conditions. The order of

the tasks as well as the task forms were counterbalanced for each sub-

group. An individual subject received either Form A or B of the Howard

mazes before the treatment condition, and the alternate form afterwards.

They received both forms of the perceptual task in counterbalanced

order before and after. The Zuckerman questionnaire was presented once,

at the end of the experiment, and the MMPI was administered at an entirely

separate time.

The prison inmates were administered the MMPI by their respective

institutions,usually as a part of a battery of diagnostic tests upon

admission. Those who did not have this information already available

in their files were then tested by the institution at their convenience

before the completion of the study.

The control subjects were administered the MMPI as the first part

of the two-part experiment in which they had agreed to participate.

Upon completing this test in a group session, they signed up for in-

dividual appointments to take Part II.

For the Stimulus Enhancement Condition, the subject was ushered

into a small room and administered the two tasks in a previously deter-

mined order. The instructions for the Howard mazes were:

"On the following pages you will find a series of mazes.
On each page you are to draw a line directly from the

"A" to the "B", staying within the boundaries. (For
Form B only, "you may choose any "B" you wish.) You
may take any path you wish as long as you go from A
to B and stay within the boundaries."

For the drawing tasks (lines or circles) the subject was blind-

folded, a pen placed in his writing hand and given the following in-


"Now I want you to draw a line (circle) three inches
in length (diameter). At this point the experimenter
places the subject's hand on the paper. Approximately
five seconds after the subject is finished, his hand
is again placed on the paper and he is told, "Now I
want you to draw another line (circle) three inches
in length (diameter)."

These instructions are repeated until 5 trials are completed.

Next, the subject was ushered into the sensory enhancement room,

where he was seated in a straight chair and told that he would be

joined by others in a few minutes. He was free to smoke, look around,

etc.,and talk as long as he did not discuss the experiment. Meanwhile,

the next two subjects were pre-tested (approximately five minutes per

subject). Then, the movie was projected for its thirty minute run.

For the post-testing, the two tasks were again administered and

the Zuckerman questionnaire was administered with the following in-


"I want you to answer this questionnaire as honestly as you
can, thinking about the way you feel about these items as of
this moment. Pick the statement that describes your prefer-
ence most accurately. Choose either A or B for each item
even if the decision is difficult to make."

In answer to questions relating to any of the tasks, the experi-

menter was instructed to give factual information (i.e., definition of a

word, repeating or rewording the instruction) but not to help the subject

make any decision in answering a test item or interpreting how the

subject was doing. The subject was to be told to do the best he could.

For the Stimulus Deprivation Group, after pre-testing, the subject

was led to another small room where he was informed about the various

equipment used and told that he would be left alone for a half hour

during which it was important that he remain as quiet and still as he

could, until the experimenter returned.

After the thirty-minute period, he was returned to the testing room

where the post-testing was completed in the same manner as the control




The mean ages for the groups were: Marion 25.4, Menard 25.2, and

SIU 20.1. The means and standard deviations for all tasks, before and

after the experimental conditions, are presented in Table 1. The Sensation

Seeking Scale showed a mean of 20.05 over all groups. The group with the

highest mean on this task was that of the Marion prisoners with a mean

score of 22.63. The other two groups achieved means of 19.55 and 19.57.

The other tasks were presented twice and the means and standard

deviations after testing are comparable in time of administration to

the Sensation Seeking Scale. The comparisons before and after, as

well as the total means, are listed in Table 1. For the maze task,

the total mean is 5.96 and as with the Zuckerman Scale, the highest

group mean belongs to the Marion prisoners at 6.65, the other two

groups receiving means of 5.76 and 5.87. Inspection of the means

before and after the experimental condition reveals a larger score on

the second administration for all groups and both conditions. This

same finding can be seen for the Line scores when before and after

means are compared. The mean Line score for all groups was 5.84

inches, and again, Marion received the largest group mean of 6.43,

with Menard prisoners at 6.37 and SIU students at 5.31. Circles were

eliminated from the analysis as no reliable method of estimating

diameters could be found for many of the data.

The Trend score is a measure of the tendency to increase or

Table I


Task Total Group Marion Menard SIU
Mean SD Mean SD Mean SD Mean SD

Maze Before 5.58 5.26 6.25 4.42 5.16 5.56 5.65 5.26
Maze After 6.34 5.39 7.05 5.23 6.40 5.90 6.08 5.28
Line Before* 5.62 2.03 5.96 2.13 6.17 2.19 5.17 1.77
Line After* 6.06 2.21 6.41 2.24 6.56 2.37 5.47 1.91
Trend Before 1.00 4.80 2.00 2.81 1.15 6.00 0.59 4.32
Trend After 0.61 3.39 1.89 2.61 0.85 3.57 0.05 3.36
SSS 20.05 5.29 22.63 4.88 19.55 5.35 19.57 5.14
Mazes 5.96 4.66 6.65 4.25 5.78 5.11 5.87 4.45
Lines 5.84 2.00 6.43 2.00 6.36 2.16 5.32 1.73
Trends 0.80 3.19 1.95 2.13 1.00 3.64 0.31 3.00


decrease the length of line over the five trials on each administration.

For each of the groups as well as the combined total group, this trend

is positive; that is, the lines increase in length with trials. This

tendency is reduced on the second administration, however, as inspection

of the before and after trends reveals. Again, the Marion group has

the highest trend mean of 1.95 as opposed to 1.00 for Menard and .31

for the SIU subjects.

The correlations between the tasks are presented in Table 2. On

each of the tasks the correlation between before and after presentations

is positive and highly significant. The mazes for the before-after

comparison ranged from .43 for SIU to .68 for Menard. Although the

forms correlated significantly (.001) for all but the Marion subjects,

these correlations were lower than the reliability estimates reported

by Howard (1961).

The Line task had higher correlations than Mazes before and after

treatment with a range from .69 for Marion prisoners to .80 for the

Menard group. These correlations were significant at the .001 level.

The Trend comparisons, however, were of a low order ranging from .10

for Menard subjects to .25 for SIU subjects. Only the SIU and the

combined total Trend correlations reached significance at the .05 level.

Of the total of 45 correlations within each institutional group

only a few were significant at the .05 level. These were Trend before

and Maze before at .49 for Marion, Lines after and Trend after at .29

(.01 level) for all groups, and .31 for Menard; Line after and Trend

mean at .29 (.01) and .28 for SIU; Trend after and Line means .19 for

the combined groups; Trend mean and Sensation Seeking Scale .22 for

combined groups and .29 for SIU.

Table 2


Variable 1 Variable 2 Total groups Marion Menard SIU

Maze Before x Maze After .53 .55* .68*** .43***
Maze Before x Line Before -.01 .10 .11 -.14
Maze Before x Line After -.03 .20 .01 -.13
Maze Before x Trend Before .12 .49* .04 .14
Maze Before x Trend After -.01 .27 -.03 -.07
Maze Before x SSS .28** .32 .23 .31*
Maze Before x Maze Mean .87*** .56*** .91*** .84***
Maze Before x Line Mean -.02 .16 .06 -.14
Maze Before x Trend Mean .08 .49* .02 .06
Maze After x Line Before .06 -.09 .04 .12
Maze After x Line After .04 .26 -.13 .07
Maze After x Trend Before .02 .19 -.03 .03
Maze After x Trend After .02 .30 -.08 .01
Maze After x SSS .21 .34 .28 .11
Maze After x Maze Mean .88*** .90*** .92*** .85***
Maze After x Line Mean .05 .10 -.05 .10
Maze After x Trend Mean .03 .30 -.07 .02
Line Before x Line After .77*** .69** .80*** .76***
Line Before x Trend Before -.01 .16 -.14 .04
Line Before x Trend After .06 .13 .03 -.08
Line Before x SSS -.05 .15 -.05 -.07
Line Before x Maze Mean .03 .00 .08 -.01
Line Before x Line Mean .93*** .91*** .94*** .93**
Line Before x Trend Mean .02 .19 -.10 .02
Line After x Trend Before .12 .03 .00 .24
Line After x Trend After .29** .30 .31* .20
Line After x SSS -.04 .29 -.23 -.10
Line After x Maze Mean .01 .26 -.07 -.04
Line After x Line Mean .94*** .92*** 95*** .94*4
Line After x Trend Mean .24** .20 .15 .28*
Trend Before x Trend After .20* .24 .10 .25*
Trend Before x SSS .25** .18 .21 .29*
Trend Before x Maze Mean .08 .37 .00 .10
Trend Before x Line Mean .06 .10 -.07 .15
Trend Before x Trend Mean .85*** .80*** .87*** .85**

Table 2 (continued)

Variable 1 Variable 2 Total Groups Marion Menard SIU

Trend After
Trend After
Trend After
Trend After
Maze Mean
Maze Mean
Line Mean

Maze Mean
Line Mean
Trend Mean
Maze Mean
Line Mean
Trend Mean
Line Mean
Trend Mean
Trend Mean

*Significant at
**Significant at
***Significant at

.05 level
.01 level
.001 level






The only other significant correlations between tasks was seen

between Mazes and the Sensation Seeking Scale. Mazes before and

Sensation Seeking Scale correlated .28 for the total groups (.01) and

.31 for SIU. The Maze mean correlated with Sensation Seeking Scale .28

(.01) for total groups and .25 for SIU. The correlations for Maze after

with Sensation Seeking Scale did not reach significance, but were also

in the range of .11 to .34.

A repeated measures Analysis of Variance was computed for the

groups by the experimental conditions for each of the tasks. The

source of variance and the F values are presented in Tables 3-6.

For each task the following main effects were compared: (a) Pri-

soners vs. SIU students, (b) Marion vs. Menard prisoners, (c) Deprivation

vs. Stimulation conditions, and for the Maze task only, (d) The order

of Forms A or B. These comparisons were done for the combined before

and after means (combined column), for the comparison of before and

after means (comparison column), for before and after separately, and

an Analysis of Covariance of before and after.

Of all the comparisons analyzed (112), only two reached significance.

One was the interaction of treatment effect by Maze form (CD) in the

after treatment comparison. This F value of 4.05 is significant at the

.05 level. The second was the means between the before and after con-

ditions on the Line task. This F value of 11.68 was significant at the

.01 level. Thus these analyses of variance did not support the hypotheses

that they were designed to test. There was no significant difference

between the groups on any of the tasks or between experimental conditions,

and the hypotheses one through four were rejected.

The data were also analyzed by a breakdown into personality variables

Table 3


Source F Values
Combined Comparison Before After Covariance

Prisoners vs SIU (A) .127 .974 .651 .031 .464
Marion vs Menard (B) .517 1.870 1.640 .000 .590
Deprivation vs
Stimulation (C) 1.100 .135 .545 1.170 .658
AC .210 1.590 1.070 .049 .748
BC .061 1.350 .635 .126 .778
Maze A or B First (D) 1.550 .223 .738 1.690 .984
AD .476 .327 .796 .098 .028
BD .774 1.040 1.650 .067 .211
CD 3.060 1.080 1.040 4.040* 2.951
ACD .994 1.640 2.300 .054 .392
BCD .654 .585 1.200 .104 .076

Error Ms 4326.500 1396.800 2793.400 2929.800 2181.700
df 108 108 108 108 107
Mean 2.480

*Significant at .05 level
**Significant at .01 level

Table 4


Source F

Prisoners vs SIU (A) .023
Marion vs Menard (B) 1.320
Deprivation vs Stimulation (C) .153
AC .051
BC .258
Error Ms 28.030

Table 5


Source F Values

Combined Comparison Before After Covariance

Prisoners vs SIU (A) 1.030 .650 1.590 .441 .228
Marion vs Menard (B) .056 .001 .048 .051 .008
Deprivation vs
Stimulation (C) .118 2.740 .061 .745 2.700
AC 1.200 1.920 .329 2.140 2.530
BC .301 2.380 .001 1.040 2.550

Error Ms 758.200 99.080 406.300 451.000 185.300
df 114 114 114 114 113
Mean 11.68**

**Significant at .01 level

Table 6


Source F Values

Combined Comparison Before After Covariance

Prisoners vs SIU (A) 1.550 1.100 1.980 .139 .005
Marion vs Menard (B) .020 1.550 .342 1.200 1.550
Deprivation vs
Stimulation (C) .006 .857 .207 .626 .819
AC .543 2.840 2.000 .367 .846
BC .729 2.300 1.960 .132 .453
Error .Ms 20.630 13.690 22.950 11.370 10.940
df 114 114 114 114 113
Mean .672

as represented by MMPI scale High or 2nd High points on either the Pd

Scale, or the Ma Scale. Tables 7, 8, 9, and 10 represent the Psychopathic

Deviate Scale (Pd) and Tables 11, 12, 13, and 14, the Hypomania Scale (Ma).

On the Psychopathic Deviate Scale analysis, only the Maze task re-

vealed some significant F values at the .05 level (Table 7). It can

be seen that these effects arise from source C and D and their inter-

actions. These are both order effects, with the position of Pd high

or second high reaching an F value of 6.13 for the comparison analysis.

The order of Form B or A of the mazes reached an F value of 4.57 when

compared after treatment. These two effects were also present in the

CD interaction for the comparison (F=5.65) column and after (F=4.52)


In the MMPI analysis of variance for the Psychopathic Deviate Scale

Lines (Table 9) the comparison mean reached significance at the .01

level with an F of 11.00 although none of the individual comparisons

in the entire table were significant. In general, the Pd scale did not

produce any meaningful relationships between tasks.

The analysis of variance by Ma 'scale scores revealed the most

significant results. Table 13, the analysis of Line scores, indicates

the main effect of A (treatment conditions) to be significant at the .01

level for combined, before, and after columns. The respective F values

are 14.46, 11.51, and 13.57. This effect is also apparent in the inter-

action with C (position of Ma being 1st or 2nd high point). The F

values for combined, before, and after columns were 16.71, 12.72, and

16.28 respectively, and all significant at the .01 level. Source B

(high or low Ma score) was significant in these three categories at the

.05 level, with F values of Combined 6.33, before 4.74, and after, 6.24.

Table 7


Source F Values
Combined Comparison Before After Covariance

Dep. vs Stim. (A) 1.230 .223 .533 1.470 .933
High vs Low Pd (B) .033 1.020 .413 .112 .619
Position 1 and 2
within high Pd (C) .331 6.130* .482 2.980 5.980*
AC .108 .990 .581 .040 .492
BC .008 1.520 .263 .473 1.260
Order of Maze Form 3.230 1.240 1.030 4.570* 3.490
AD 3.320 .382 1.630 3.670 2.090
BD .154 .053 .202 .055 .000
CD 1.170 5.650* .044 4.520 6.850*
ACD 1.270 .039 1.140 .808 .154
BCD .309 .978 .000 .998 1.310

Error Ms 4197.200 1301.700 2814.500 2684.600 1957.900
df 108 108 108 108 107
Mean 2.66

*Significant at the .05 level

Table 8


Source F

Dep. vs Stim. (A) .124
High vs Low Pd Scale .064
Position 1 and 2 within high Pd (C) 1.090
AC .176
BC 1.320
Error Ms 29.080
df 114

Table 9


Source F Values

Combined Comparison Before After Covariance

Dep. vs. Stim. (A) .005 1.580 1.490 .232 1.420
High vs. Low Pd (B) .193 .646 .520 .016 .419
Position 1 and 2 within
high Pd (C) .207 .013 .236 .138 .000
AC 2.720 2.110 1.190 3.880 3.040
BC 2.970 .078 3.200 2.130 .022

Error Ms 768.800 105.200 401.100 473.000 200.200
df 114 114 114 114 113
Mean 11.000"**

***Significant at .01 level

Table 10


Source F Values

Combined Comparison Before After Covariance

Dep. vs. Stim. (A) .082 1.890 .889 .626 .124
High vs. Low Pd (B) 1.450 1.900 2.440 .007 .064
Position I and 2 within
high Pd .001 2.110 .602 1.320 1.090
AC .001 1.430 .449 .808 .176
BC .100 1.710 .251 1.720 1.320
Error Ms 20.880 13.790 23.140 11.530 29.080
df 114 114 114 114 114
Mean .667

Table 11


Source F Values

Combined Comparison Before After Covariance

Dep. vs Stim. (A) 2.080 .352 2.450 .923 .026
High vs Low Ma Scale (B)1.740 1.250 2.940 .359 .120
Position 1 or 2 within .110 .223 .004 .265 .319
high Ma (C)
AC .190 .095 .054 .277 .224
BC .483 1.160 .007 1.250 1.590
Order of Maze Form (D) 3.260 2,490 5.620* .638 .258
AD 4.550* .699 5.260* 2.080 .073
BD .381 3,710 2.230 .157 1.870
CD .256 1.190 .008 .927 1.400
ACD .000 .501 .117 .121 .381
BCD .016 1.390 .221 .457 1.160

Error Ms 4105.400 1279,200 2636.800 2749.900 2009.100
df 108 108 108 108 107
Mean 2.71

*Significant at .05 level

Table 12


Source F

Deprivation vs Stimulus (A) .032
High vs Low Ma Scale (B) .901
Position 1 or 2 within high Ma (C) 2.460
AC .498
BC 2.520

Error Ma 28.680
df 114

Table 13


Source F Values
Combined Comparison Before After Covariance

Dep. vs Stim. (A) 14.46** .518 11.51** 13.57** 2.78
High vs Low Ma Scale(B) 6.33* .424 4.74* 6.24* 1.67
Position 1 or 2 within .024 .022 .028 .015 .000
high Ma (C)
AC 16,71** .972 12.72** 16.28** 3.87
BC .032 .029 .056 .010 .011

Error Ms 694.7 106.8 374.3 427.2 198.0
df 114 114 114 114 113
Mean 10,83*

Significant at .05 level
** Significant at .01 level

Table 14


Source F Values

Combined Comparison Before After Covariance

Dep. vs Stim. (A) .419 5.039* .629 5.81* 6.60*
High vs Low Ma Scale (B) 1.24 .036 .412 1.69 1.18
Position 1 or 2 within .781 .824 1.19 .010 .009
high M (C)
AC 2.20 .153 .610 3.87 2.48
BC 2.07 .055 .177 .575 .022

Error Ms 20.92 13.90 23.29 11.48 11.10
df 114 114 114 114 113
Mean .662

*Significant at .05 level

These results imply that the groups showed significant differences

before treatment as well as after and that the before-after comparison

was not affected. This finding of initial differences between groups

given either stimulation or deprivation was not found in either the Pd

or institutional group analyses.

*Other effects of the Ma analysis revealed significant F values at

the .05 level for experimental conditions on the Trend task. This was

noted in the comparison F (5.039), the before F (5.81) and the covari-

ance (6.60). The trend here was a treatment effect of lowering the

overall increase rate of length of line. The groups started out signi-

ficantly different and then became similar after treatment.

The maze task showed a significant order effect with the A or B

form being different before treatment for the stimulation and deprivation


Ma as a high profile point does not necessarily operate to differ-

entiate groups in the same manner as a correlation between Ma and Pd

scales with all the tasks, before, after, and combined are presented

in Table 15. The highest correlation on this table, that between Ma

and Trend after for the Marion group is -.40 which is not significant at

the .05 level. The MMPI data with the prisoner and student groups are

presented in Table 16, which contains the means and standard deviations

of all groups on each of the scales. The SIU group fell within the

normal range with high points on scales Sc (Schizophrenia) and Ma. The

Menard group showed higher elevations for scales Pd and Ma, with Pd

reaching a mean of 71.27 and Ma, 64.35. The Marion group had an

extremely deviant mean profile with five T scores above 70; Pd at 87.42;

Sc at 86.95; Pt at 81.95; Hs at 70.79; and Ma at 77.48.

Table 15



Pd or Ma Task Total Groups Marion Menard SIU

Ma x Maze Before -.60 -.10 .22 -.01
Ma x Maze After .07 .06 .19 -.06
Ma x Line Before -.02 -.13 -.12 .02
Ma x Line After .00 -.31 -.16 .07
Ma x Trend Before -.01 .12 -.19 .04
Ma x Trend After -.03 -.40 -.19 .02
Ma x SSS .16 -.05 .12 .11
Ma x Maze Mean .10 .01 .23 -.04
Ma x Line Mean -.01 -.24 .15 .05
Ma x Trend Mean -.02 .16 -.25 .04
Pd x Maze Before .07 -.16 .13 .09
Pd x Maze After .11 -.12 .00 .13
Pd x Line Before .15 -.31 -.08 .19
Pd x Line After .15 -.33 -.21 .12
Pd x Trend Before .05 -.20 .03 -.04
Pd x Trend After .00 .03 -.26 .19
Pd x SSS .21 .19 .06 .13
Pd x Maze Mean .10 -.01 .07 .13
Pd x Line Mean .16 -.35 -.15 .17
Pd x Trend Mean .04 -.20 -.11 -.14

Table 16


Scale Total Group Marion Menard SIU

Mean Mean Mean Mean

L 48 45 50 48
F 59 62 58 59
K 54 52 57 52
Hs 57 71 55 54
D 58 60 55 58
Hy 58 60 58 58
Pd 67 87 71 59
Mf 58 62 56 59
Pa 56 57 57 56
Pt 63 82 57 60
Sc 66 87 60 62
Si 50 48 50 50
Ma 66 77 64 63

Because of the unusual deviant elevation of this mean profile, the

Marion group was scanned for any questionable validity scales. Three

of the 19 subjects had F scores over 70. An adjusted mean profile that

is presented in Table 17 in the last column was calculated with these

three subjects omitted. The mean differences were negligible. The

profiles of the Marion subjects were then compared to the profiles of

the general population from which they were drawn (maximum security

inmates at the Marion Federal Penitentiary). A comparison group of

minimum security prisoners at the same institution is also included

and all groups are presented in Table 17. Inspection of this table

reveals a similar profile pattern for all groups differing in elevation

with the minimum security group most normal, the general population of

maximum security prisoners moderately deviant (3 T scores over 70) and

the selected subjects (by age and education) the most deviant group.

Table 18 shows a chi square analysis of the high-low breakdown

of Ma and Pd scales according to institution. The X2 is 40.19 and

significant at the .01 level. Inspection of cells reveals that the

SIU students are predominately low Ma and low Pd, while only three

subjects had both Ma and Pd high points. The Menard group had pre-

dominately high Pd scales, whether or not Ma was also high.

The similarity of Ma in the prisoner groups may be seen by the

group profiles, Table 16, where Marion peaks higher on scales Sc and Pa,

although Ma is also very high. Thus, the population differences between

the three institutions are largely apparent in the elevation of Pd, Ma,

as well as other MMPI scales.

The Sensation Seeking Scale was also submitted to further analysis.

A chi square for high (scores over 16) versus low by institution is

Table 17


Scale Maximum Minimum Experimental Adjusted
Security Security Subjects Profile

L 51 53 45 46
F 67 58 62 59
K 53 55 52 53
Hs 58 60 71 67
D 66 63 60 61
Hy 62 59 60 57
Pd 77 68 87 86
Mf 58 55 62 61
Pa 61 56 57 53
Pt 63 58 82 79
Sc 70 60 87 81
Ma 67 59 77 76
Si 54 52 48 46

Table 18


High Point Patterns
Pd (low) Pd High Ma High Ma High TOTAL
Ma (low) Ma Low Pd Low Pd High

Marion 6 7 3 3 19
Menard 3 18 6 13 40
SIU 30 6 19 3 58

TOTAL 39 31 28 19

= 40.19 P .01

Table 19


Pd Low Pd High

Marion 9 10
Menard 9 31
SIU 49 9

*X2 = 38.07 P .01

Table 20


Ma Low Ma High

Marion 13 6
Menard 21 19
SIU 49 25

*X2 = 1.37 P .10

Table 21


Marion Menard SIU

Low 1 (5.3%) 10 (25.0%) 12 (19.7%)
High 18 (94.7%) 30 (75.0%) 49 (80.3%)

*X2 = 3.26, P

Table 22


Item No. S-S Alternative Marion % Menard % SIU % X2 P

2 A 100 50 55 14.87 .01
6 A 63.2 25.0 14.8 17.74 .01
12 B 84.2 77.5 57.4 7.15 .05
28 B 52.6 27.5 55.7 8.18 .05
29 B 68.4 35.0 37.7 6.72 .05

presented in Table 21. This X2 of 3.26 is not significant. However,

the percentage of sensation seekers in all groups is very high with

94.7% of the Marion group, 75% of Menard, and 80% of SIU. An item

analysis on the entire test (Appendix B) revealed only five items that

discriminate between groups significantly at the .05 level or beyond.

These items are listed in Table 21. In each case (except item 28) the

Marion group showed the highest percentage answering in the sensation

seeking direction. As noted before, Marion had the highest means on

this task. So, although all analyses on the Sensation Seeking Scale

were non-significant, trends for the most deviant group (Marion) to be

the highest on this task were present in the data.

In general, the analyses of all the data, while they did not support

the stated hypotheses, did reveal that the Howard Maze Test, the Zuckerman

Sensation Seeking Scale, and the Lines means and Trends scores tended to

follow a pattern in the direction of greater sensation seeking for the

most socially deviant group, the federal maximum security prisoners at

Marion. The State penitentiary prisoners (Menard) and the SIU college

students appeared to be more similar than different on these measures.

The tasks, while having only a few low significant correlations on these

groups, did seem to operate in the same direction, with sensation seeking

the predominant mode for the Zuckerman scale, and the reduction of lines

and trends, a phenomenon that Petrie notes as belonging to sensation

seekers as well as delinquents. The conditions did not have any con-

sistent effect other than some sporadic order and randomization effects.

The MMPI analyses were not much more successful in differentiating

the tasks and their effects. The Lines task showed some highly significant

F values which indicated that the groups were different before treatment


on the Ma breakdown. F scores on the other tasks are best interpreted

as chance significant effects. There was no particular pattern or con-

sistency to them and,in view of the large number of F values tested, the

probability is high that some false significant scores might occur. For

each of the three breakdowns of analyses of variance there were 112

separate F values.



The present study tested the following hypotheses: (a) Criminals

will be higher on sensation seeking tasks than will their noncriminal

controls. The three tasks selected were the Zuckerman Sensation Seek-

ing Scale, the Howard Maze task, and a line drawing task, comparable to

Petrie's kinesthetic task. (b) Sensory restriction conditions will in-

crease the amount of sensation seeking on these tasks for all subjects.

(c) Restriction of sensory stimulation will have a greater effect on

criminals than on controls. (d) There will be no differences in dir-

ection of results for the various tasks, since sensation seeking is

considered a general factor. (e) High MMPI Ma and Pd scale scores will

characterize sensation seekers across all experimental groups.

The testing of the first hypothesis did not yield significant dif-

ferences, although the federal prisoners showed trends in the expected

direction on both the Zuckerman Scale and the Mazes. The second and

third hypotheses were rejected as there were no meaningful effects for

any group as a result of deprivation or stimulation conditions. Hypo-

thesis(d)was partially supported by the data. There was a low but sig-

nificant correlation (+.28) between the means of the Howard Mazes and

the SSS. This indicated that while the two measures are significantly

related, this relationship accounts for only a small portion of the

variance. Also, a trend pattern emerged in that all groups rose on

their mean scores, indicating similar change patterns in the sensation

seeking measures by retesting. Failure to support hypothesis (d) was

seen in the nonsignificant correlations between the other tasks. The

fifth hypothesis (e) was not supported. The breakdown by MMPI scales

Pd and Ma did not appreciably change the results or aid in identifying

sensation seekers. Neither scale correlated with the sensation seeking


The failure to support the hypotheses and the partial support for

hypothesis (d) will be discussed in terms of subject population, the

relationship of antisocial behavior and sensation seeking, adequacy of

the sensation seeking tasks, the generality of the characteristics of

sensation seeking, the limitations of measuring stimulus conditions,

and the relationship of anxiety to sensation seeking in the manipulation

of levels of arousal.

Criminals as a Subject Population

The rationale for the selection of criminals as a group represen-

tative of psychopaths was based on the demonstrated behavior that they

are capable of severe antisocial acts in the face of great risk (being

caught and incarcerated). This group is separated from the noncriminal

college students, who may have the potential for performing illegal,

felonious acts but have either the controls to contain the impulses to

do so or have thus far managed to avoid being caught. Age and educational

factors are often considered the most important factors in impulse

control. This was controlled in the current study by a subject require-

ment of being enrolled in college level courses and an upper age limit

of 30 years.

Using prisoners as subjects has often presented difficulties in

interpreting the data. Morello (1959 for example, suggests that maxi-

mum security penitentiary inmates are too sophisticated for reliable and

valid measurement by a questionnaire type of psychological test. Rcttig

(1964) has found that they become cautious on questionnaires more read-

ily than when not in confinement. On maze type tasks (Docter and Winder,

1954) Rankin and Wykoff, 1964), the behavior of delinquents has revealed

their freedom to disobey instructions and to interpret rules and barriers

with less rigidity. This behavior also seems to be the case with extro-

verts and other impulsive subjects in sensory deprivation studies (Tranel,

1962). One explanation for their seemingly greater tolerance in some of

the experiments appears to be due to the fact that they move around more,

fall asleep, etc.,and thereby are not deprived in the sense of the sub-

ject who strictly adheres to instructions. Although this attitude of

looseness of adherence to the rules of the experiment was not particu-

larly noted in the present study, the sensation seeking scale in parti-

cular seemed to bring questions from the subjects that indicated that

they were not comfortable with the decisions that they were asked to

make on this questionnaire. The pairs of items were not always seen as

opposing choices. The items posing most difficulty were numbers 4, 5,

13, and 22 (over 5 queries each at Menard) and to a lesser degree, items

11, 25, and 34. One S remarked after long deliberation to item 22, "If

I can't answer both I ain't answering either."

The selection factors of age and education restricted the range of

differences between the groups on the sensation seeking tasks. Kish and

Busse (1968) found that both these factors were significant in contribut-

ing to Sensation Seeking Scale variance. Volkman criticizes most MMPI

studies used to differentiate personality patterns of delinquent and

nondelinquent subjects for similar reasons. He noted that significance

is not reached if these factors of age and education are accounted for;

however, in his own study, a small N may have contributed to lack of

results (Waldo and Dinitz, 1967).

The chi square analysis in the present study revealed a significant

separation of the prisoner and nonprisoner groups according to the scores

on the Pd scale of the MMPI. The total profile patterns further empha-

sized the similarity of the results with those of other researchers

(Dahlstrom and Welch, 1960). Prisoners who act out in confinement, as

determined by such measures as disciplinary reports, tend to have more

deviant MMPI profiles, notably a rise in F and Sc as well as Pd and Ma

(Blackburn, 1969; Craddick, 1962). High F scores are not considered in-

valid in these particular groups as it is frequently encountered and

noted to have clinical meaning (Mack, 1969).

The differences between the Marion and Menard prisoner groups in

the present study may be seen as one of degree of deviancy. Comparison

of the mean profiles shows Marion prisoners' profiles to be similar to

those at Menard, but more highly elevated, particularly on the critical

scales, Pd, Pt, Sc, and Ma. In order to qualify for the population of

maximum security at the Marion Federal Penitentiary, an inmate must have

been transferred (usually because of bad behavior) from another federal

institution. The Menard population, while accepting prisoners with

serious crimes (Appendix G lists the crimes committed by all subjects),

does not draw its population from disciplinary cases of other institu-

tions, but rather accepts direct court commitments. The crime breakdown

in Appendix G suggested that there is no particular difference in sever-

ity or type of crime committed.

Although, in general, prisoners and delinquents demonstrate deviant

behaviors in test taking and casualness about instructions, it does not

seem that this in itself constitutes a basis to ignore the differences

between the criminal groups and their college controls. In the present

study control of age and education variables may have reduced variabil-

ity between prisoners and controls found in other studies. The differ-

ences within prison subject groups may provide more useful bases for

future investigations of sensation seeking.

Antisocial Behavior and Sensation Seeking

In a few studies attempting to relate psychopathy to sensation

seeking, Blackburn (1969), Leubeck (1967), and Petrie (1967) have used

instruments purporting to measure sensation seeking behaviors as depen-

dent variables. Luebeck's results were nonsignificant, Petrie found a

high relationship between the reducing style of perception and delin-

quency, and Blackburn found only one MMZPI Scale (Ma) to correlate with

psychopathy. Notably in the latter two studies, however, other findings

concerning the reactions of the psychopaths were of relevance. Petrie

found that about 14% of the delinquents were neither reducers or aug-

menters, but stimulus governed. That is, they perceptually exaggerated

the normal contrast effect of calling the small block larger and the

larger block smaller than if the comparisons were not made. Thus certain

perceptual immaturities may exist in antisocially prone individuals in

addition to a tendency to reduce stimuli. In Blackburn's study, corre-

lations with MMPI Pa, Sc L, L F, and K Scales with sensation seeking sug-

gested to him that some psychotic type patterns were contributing to the

higher SSS scores. Similarly Petrie has suggested that sensation seekers

in extreme need states suffer from a feeling of unreality where there is

not enough structure to the environment. Behavioral controls then may

be tenuous because of impulsivity or because of a more general reality


The deviant profile of the Marion subjects in the present study may

represent an additional variance contributing to sensation seeking scores,

other than the impulsive, hostile, antisocial tendencies one typically

associates with severe antisocial behavior. There may exist in these

subjects a difficulty in reality control that goes beyond a difficulty

with impulse control.

Adequacy of Sensation Seeking Tasks

The particular instruments used in this study merit discussion in

light of the failure to support almost all the hypotheses. The diffi-

culty with some items on the Sensation Seeking Scale has already been

mentioned. The other tasks did not present any administration difficul-

ties. Both Lines and Mazes were dependent on motor production rather

than verbal expression. In each case, the task took only a few minutes.

Both Mazes and Lines exhibited good reliability. The correlations be-

tween before and after correlated .53 for the total group, which is sig-

nificant at the .001 level. This relationship has, of course, been

lowered by any differences which may have resulted from treatment differ-

ences, Domino (1965) reports test-retest reliability at .76 after one

week, and .71 after six months for Form A of the Howard Mazes. The Lines

task before and after scores correlated .77, significant at the .001

level. There is no comparable test-retest measure for this task as it

was especially designed for this study as an adaptation of Petrie's

kinesthetic task, which has split-half reliabilities of over .90. The

Sensation Seeking Scale was administered one time in this study, after

treatment. This scale has been noted to have variable regional norms

(Zuckerman, 1969b)and reliability and validity data is still being

collected. Zuckerman is now in the process of developing a revised

form (Zuckerman, 1969a).

The Trend scores showed a low before-after correlation of .20, sig-

nificant at the .05 level. Their correlation with the Sensation Seeking

Scale was higher, .25, significant at the .01 level. The Trend score,

thus, does not appear to be a very reliable measure, and the adaptation

effect that this was suggested to represent is not consistent.

This task was dependent on the active production of the subject,

rather than an estimate from presented stimuli as was Petrie's task.

Perhaps exaggeration of stimuli produced is not affected by sensation

seeking to the degree that perception of stimuli by the environment is.

According to Petrie's theoretical arguments, however, the perceptual

styles of reducers and augmenters should be capable of shifting from

sensory modes as well as from task to task.

Thus, the present study confirms that the line task has satisfac-

tory reliability. Reliability of the mazes was considerably lower in

this study than in simple reliability studies. The reliability of the

trend scores was not satisfactory. The measures used were the most

frequently found in the literature and seem to represent the best avail-

able at the present time. The current study indicates that there is a

continuing need to develop and validate better measures of sensation


Sensation Seeking as a General Factor

The issue of the generality of sensation seeking was supported to

some degree by the results of this study (Hypothesis 4). The means of

the two tasks that were administered twice, Mazes and Lines, both in-

creased slightly on the second administration. The Maze mean and Sen-

sation Seeking Scale correlated .28 (.01 level), in contradiction to

Zuckerman's findings of a -.19 correlation (1964). However, his N was

considerably smaller, 40 as opposed to 120 on the present study. An

item analysis of the Zuckerman scale on the present study indicated

very few items (5) that significantly differentiated between the subject

groups (Appendix B). This instrument is not very useful as a discrimin-

ator of prisoners, as it now exists. Many items, however, showed some

discriminant ability at the .10 level and the Marion group in almost all

cases chose the sensation seeking alternatives on these items.

Since the present data were gathered, Zuckerman has revised his

scale to a 72 item form with one general scale and four factor scales

(Zuckerman, 1969b). He has identified the factors as (1) thrill and

adventure seeking, (2) experience seeking, (3) disinhibition, and (4)

boredom susceptibility. A direction for further research would be to

test the ability of the new scale to discriminate between different

aspects of sensation seeking preferences for personality groups. The

present Form II reveals conflicting results particularly in its rela-

tionship to anxiety and to other sensation seeking tasks such as the

Howard mazes (Zuckerman and Link, 1968; Zuckerman et. al., 1964).

In the current study, the low but significant correlation between

Howard's mazes and the Sensation Seeking Scale, suggests the possibility

of some commonality of sensation seeking in different forms. The issue

of more specific sensation seeking factors is an avenue of research that

needs further exploration. Maddi (1965) has tentatively explored this

alternative with the use of TAT protocols to elicit productions of stories

high in novelty, curiosity, or the desire for novelty. He found that the

person who expresses considerable desire for novelty is unlikely to com-

pose unusual or curiosity-laden stories. This raises the question of

whether or not the desire for novelty has a close relationship with the

need for variety. Skrzypek (1969),whose study is most closely related

to the present one,found a significant correlation between psychopathy

and preference for complexity and novelty. However, his deprivation

conditions did not induce a rise in the preference on retest of the

tasks, so the results can be interpreted as a desire for complexity and

novelty among psychopaths that may or may not be based on a state of need.

Future factor analytic studies of the Zuckerman scale may provide

more information on the generality of the concept of sensation seeking.

Comparison between various types of measures as they become more refined

in predicting sensation seeking will also shed light on this issue. For

the present, the generality of sensation seeking remains uncertain.

Limitations of the Measurement of Stimulation Conditions

The lack of effect of stimulation differences on the performance of

the subjects in the present study may be interpreted as being due to

either the lack of direct relationship between need states of stimulation

and the particular tasks used, or as an alternative, a lack of adequate

manipulation of the subject's need for stimulation. The experimental

manipulation of stimulus conditions offers many problems. Most studies

have dealt with only one direction,deprivation. A suitable control for

deprivation studies is difficult to find and most studies either ignore

using control subjects or use conditions that amount to certain degrees

of deprivation themselves, such as confinement in small rooms, alone,

with only magazines for reading, or a dull, unstimulating task to per-

form (Zuckerman, 1969a). Of the few studies that used control conditions

for stimulation, only Skrzypek (1969) found significant results on his

tasks. His stimulation conditions consisted of a 30-minute, boring task,

with the final 10 minutes involving an intensely stressful concentration

task in which the subjects were baited with monetary rewards, and also

told, "others almost always get it right." Whether the results were

due to the high level of stimulation, or the effects of an emotionally

arousing state of anxiety,is impossible to determine from his study.

His neurotics did decrease their complexity preference scores, but this

may have been due to other heightened stimulation or to anxiety, a very

specific kind of stimulation.

Zuckerman, et. al. (1969),did have an extremely contrasting stimu-

lation-deprivation effect in his study. He did not find any differences

between his groups on the SSS even though the stimulus period included

strob lights, multimedia films and slides and multiple recordings. He

concluded that even his amount of stimulation may not have been suffi-

cient for the present young generation which is accustomed to very loud

sounds, bright lights and colors. One possibility for his lack of re-

sults may have been that none of his stimulation was meaningful (pur-

posely on his part), According to the theoretical position of Fiske

and Maddi (1961), meaning is an important aspect of the impact of stim-

ulation beyond the very basic stimulation levels. The present study

attempted to include meaningful stimulation by means of action movies

with a theme, whose content was felt to have no anxiety producing

connotations, although the degree of meaning for individuals may have

varied. Perhaps stimulation through visual and auditory modes only

may be more easily defended against or adapted to than more physical

limitations. One must attempt to involve more active stimulation,

motorically, or emotionally, in order to measure differences on the

type of tasks presented.

In addition to contrast between deprivation and stimulation, the

duration of stimulation or deprivation conditions may be seen as a com-

plicating factor in the interpretation of the results. Although most

deprivation studies are longer than the present study, Skrzypek (1969)

was able to produce significant results with 40 minutes. Also, there

is evidence that confounding factors become involved at different periods

of deprivation. Short term studies, under one day, tend to give differ-

ent results than long term studies. Even within the first few minutes

of deprivation, effects can be noted, but these effects apparently can

change with adaptation to the situation as the duration is lengthened

(Zubek, 1969b). EEG measures during the first hour of total isolation

showed repeated periods of low, slow waves. These frequently appeared

within the first 10 minutes (van Wulfften Palthe, 1957). Also, decrease

in skin conductance and heart rate and a rise in pain sensitivity have

been found in deprived subjects within the first hour, only to rise

again in fluctuations later (Zubek, 1969b).

Because of the conflicting findings between studies with varying

lengths of deprivation, it seems necessary that confinement times be

systematically varied for individual subjects as well as the intensity

of the experience. Petrie (1967) cautions that the extremes of sensa-

tions may give rise to a defensive reaction of the body that causes the

augmenter to reduce when a certain intolerable amount of sensation is

bombarding the individual. She has evidence that the bombardment of

sound causes the augmenter to reduce, but has no significant effect on

the reducer. This is also supported by Stennett (1957) who found an

inverted U-shaped relationship between level of arousal and performance.

At the very highest stimulus levels, performance went down, whether

measured by EMG or palmar conductance. While this relationship may hold

for performance with regard to the effect of stimulation on an individual,

the optimal point on this continuum should, according to Fiske and Maddi's

theory, differ for the different, momentary need levels of stimulation.

The unanswered question posed by the present study,as well as the others

that did not vary the intensity and duration of the stimulus conditions,

is where in relation to his optimal level of stimulation is each subject

when placed under these specific conditions? It is perhaps more impor-

tant for the present state of knowledge on this matter to investigate

small numbers of individuals under varying conditions than to ascertain

means for large groups at any one level of intensity or duration.

Relationship of Anxiety to Sensation Seeking

Although this study did not purport to investigate the relationship

of anxiety to psychopathy and sensation seeking, it is perhaps necessary

to comment on this issue since anxiety appears to be difficult to separ-

ate from stimulation as it is often measured. Helson (1964) proposes

that adaptation may cause stimuli to change from pleasant to indifferent

to unpleasant,or vice versa, so that with continued repetition there is

a reversal of affect. Thus, anxiety or emotional reaction to a situation

as well as the need for stimulation may change with changing conditions.

Anxiety has already been shown to be influenced by sensory deprivation

by Zuckerman, et. al. (1969). He found that there was more physiologi-

cal contrast evident between his experimental conditions than ques-

tionnaire differences in his study. The affect associated with the

deprivation effect in the present study has been observed to vary con-

siderably. Comments ranged from "nice sleep," "good rest," "I enjoyed

it," to "thank goodness it's over," "it seemed like ages," For some,

then, 30 minutes was already stressful, for others,a much longer period

might remain pleasant. Actual situational anxiety seemed minimal, how-

ever, as the subjects were told beforehand the length of time they would

stay and all that was expected of them. Further studies dealing with the

issue of sensation seeking should either attempt to include a measure of

affect as well as stimulation, or eliminate the possibility of a con-

founding of the two variables.


In the present study, the populations may have been too close in

demographic characteristics to pick up enough variance to reach signi-

ficance for the size of the N. Originally, prisoners were to be com-

pared to college students, but due to large differences in prisoner

groups, this main group was subdivided in the analysis. Thus far, no

study controlling for age and education has revealed differences in

deprivation versus arousal for psychopaths. Skrzypek's study, using

extreme sensation avoiders (neurotic delinquents) as controls,found

differences only related to anxiety (Skrzypek, 1969).

Control conditions for deprivation represent a methodological

problem which has not yet been adequately resolved. It must include

meaningful stimulation controls of varying intensity and duration,

unrelated to stress of anxiety or other negative emotional content.

Although the present study attempted to use all but the varying stim-

ulation controls, results were not significant. However, the direction

of changes in means of the groups indicates that the trend may exist in

a more powerful test of the hypotheses. Thus far, the question of the

relationship of psychopathy to sensation seeking is still unclear,

Quay's theory of pathological sensation seeking in psychopaths remains

equivocally supported by the present state of knowledge concerning the

basic reactivity of psychopaths to stimulation as well as their adapta-

tion to stimuli. It is felt, however, that an adequate test of this

theory is still possible although some more basic questions relating

to sensory deprivation conditions and instruments used to measure sensa-

tion seeking must still be resolved.

The difficulties encountered in the present study in terms of ex-

perimental conditions, measurement instruments,and subject population

are all reflective of problems in the present stage of research in the

area. The contribution of this study may by seen in its presentation

of negative results with the most promising available research tools,

Correcting the shortcomings of these tools is a necessary step to find-

ing a meaningful answer to the question of the relationship between sen-

sation seeking and criminality.



The theory postulated by Fiske and Maddi states that the optimal

stimulation level differs for individuals according to their basic

differences in need for stimulation as well as their momentary arousal

states. There is a growing body of evidence that criminals and other

hostile and impulsive groups seek more stimulation than nonoffenders.

Quay has proposed that the psychopath is at the pathological extreme

of the continuum.

Sensation seeking as a personality variable contributing to severe

anti-social behavior was investigated by the use of environmental manipu-

lation of available stimulus input for prisoners and college student

controls. Two tasks purporting to measure sensation seeking were

administered before and after treatment, and a third questionnaire

measure, afterwards.

It was proposed that (a) Criminals would be higher than noncriminals

on all sensation seeking tasks, (b) Deprivation of stimulation would in-

crease scores in the direction of greater need, (c) The three types of

sensation seeking tasks, perceptual, motor performance, and questionnaire

would operate in a similar manner for both groups, implying a general sen-

sation seeking factor, and (d) Analysis by MMPI Pd and Ma Scales would

yield similar results for the foregoing hypotheses. This analysis was

included to test the possible difference in use of a personality trait

rather than actual demonstrated antisocial behavior (criminality) as

an independent variable.

The subjects were 59 prisoners from two maximum security prisons

in Illinois, 19 from a federal penitentiary and 40 from a state insti-

tution. Control subjects were 61 college introductory psychology

students. All subjects were under 30 years of age and currently enrolled

in college courses.

The subjects were randomly assigned to either the sensory restric-

tion or sensory enhancement groups. They were tested before and after

treatment on two tasks, the order counterbalanced. The perceptual task

was an adaptation of Petrie's kinesthetic task in which the subject

was asked to draw a three inch line while blindfolded, and given five

trials at each testing. The motor performance task was Howard's Maze

Task, a paper and pencil maze test with alternate paths to a goal,

scored according to the number of changes in paths taken on five

successive trials. The questionnaire administered upon completion of

the experiment was Zuckerman's Sensation Seeking Scale, a forced choice

test consisting of 34 pairs of preferences for activities, one of each

pair being the sensation seeking alternative.

Sensory restriction consisted of 30 minutes in a darkened, quiet

room, reclining on a padded lounge chair, while wearing earplugs, head-

phones, a blindfold, and cloth wrapped around the hands. In the stimulus

enhancement condition, subjects in groups of three watched Laurel and

Hardy comedy films.

A repeated measures Analysis of Variance yielded no significant

results for hypotheses (a), (b), and (d) with the exception of some

few differences between treatment groups before testing in the Ma

analysis only, suggesting a chance significant difference due to

randomization effects. Hypothesis (c) was partially supported in that

the means rose for groups after treatment, and also there was a signi-

ficant correlation between the Howard mazes and the Sensation Seeking


Comparison of prisoner groups revealed that the federal penitentiary

prisoners were more deviant than the state penitentiary prisoners with

a mean MMPI profile of five T scores elevated beyond T 70. This federal

prisoner group also showed higher means on all sensation seeking tasks.

A chi square analysis revealed significantly higher proportion of high

point Ma scale scores among the group. The results on population

differences suggested that sensation seeking may have a relationship

to the degree of deviancy of those who commit antisocial acts. The

failure to support the presented hypotheses were discussed in terms

of the relative immaturity of the measurement instruments, their

generality, and the difficulties in determining the intensity and

duration of external stimulus conditions in relation to an individual's

optimal stimulation level.


Berlyne, D. E. Curiosity and exploration. Science, 1966, 153, 25-33.

Berlyne, D. E. & Lewis, J. Effects of heightened arousal on human
exploratory behavior. Canadian Journal of Psychology, 1963, 17,

Berggren, R. Autokinesis: A sensory deprivation interpretation.
Paper presented at the Eastern Psychological Association, 1968.

Blackburn, R. Sensation seeking, impulsivity, and psychopathic per-
sonality. Journal of Consulting and Clinical Psychology, 1969,
33, 571-574.

Brownfield, C. A. Optimal stimulation levels of normal and disturbed
subjects in sensory deprivation. Psychologia, 1966, 9, 27-38.

Cleckley, H. M. Psychopathic states. In S. Arieti (Ed.) American
Handbook of Psychiatry. New York: Basic Books, 1959, 1, 567-588.

Craddick, R. A. Selection of psychopathic from non-psychopathic prisoners
within a Canadian prison. Psychological Reports, 1962, 10, 495-499.

Dahlstrom, W. G. & Welsh, G. S. An MMPI Handbook: A guide to use in
clinical practice and research, University of Minnesota Press,
Minnesota, 1960.

Docter, R. F. & Winder, C. L. Delinquent versus nondelinquent perform-
ance on the Porteus Qualitative Maze Test. Journal of Consulting
Psychology, 1954, 13, 71-73.

Domino, G. A validation of Howard's test of change-seeking behavior.
Educational and Psychological Measurement, 1965, 15, 1073-1074.

Fairweather, G. W. The effect of selected incentive conditions on the
performance of psychopathic, neurotic, and normal criminals in a
serial rote learning situation. Unpublished doctoral dissertation,
University of Illinois, 1953.

Farley, F. H. & Farley, S. V. Extraversion and stimulus-seeking motiva-
tion. Journal of Consulting Psychology, 1967, 31, 215-216.

Fiske, D. W. Effects of monotonous and restricted stimulation. In
Fiske, D. W. & Maddi, S. R. Functions of Varied Experience.
Homewood, Illinois; Dorsey Press, 1961, 106-144.

Fiske, D. W. & Maddi, S. R. A conceptual framework. In Fiske, D. W.,
& Maddi S. R., (Ed.) Functions of Varied Experience. Homewood,
Illinois: Dorsey Press, 1961, 11-56.

Fooks, G. & Thomas, R. R. Differential qualitative performance of
delinquents on the Porteus Maze. Journal of Consulting Psychology,
1957, 16, 351-353.

Fry, F. D. A normative study of the reactions manifested by college
students and by state prison inmates in response to the MMPI,
Rosenzweig Picture-Frustration Study, and the TAT. Journal of
Psychology, 1952, 34, 27-30.

Garlington, W. K. & Shimota, H. E. The Change Seeker Index: A measure
of the need for variable stimulus imput. Psychological Reports,
1964, 14, 919-924.

Gibbens, T. C. N. The Porteus Maze Test and delinquency. British
Journal of Educational Psychology, 1958, 28, 209-216.

Glanzer, M. Stimulus satiation: An explanation of spontaneous alter-
nation and related phenomena. Psychological Review, 1953, 60,

Glaudin, V. Jr. The speed of visual recognition of primitive, social,
and neutral words by criminal psychopathic, criminal normal, and
hospitalized psychoneurotic subjects. Unpublished doctoral disser-
tation, University of Illinois, 1954.

Gurvitz, M. Intelligence factors in psychopathic personality. Journal
of Clinical Psychology, 1947, 3, 194-196.

Hebb, D. 0. The Organization of Behavior. New York: Wiley, 1949.

Helson, H. Adaptation-Level Theory. New York: Harper & Row, 1964.

Hocking, J. & Robertson, M. Sensation seeking scale as a predictor of
need for stimulation during sensory restriction. Journal of
Counseling and Clinical Psychology, 1969, 33, 367-369.

Howard, K. I. A test of stimulus-seeking behavior. Perceptual and Motor
Skills, 1961, 13, 416.

Howard, K. I. & Diesenhaus, H. I. Handbook for the Maze Test. Institute
for Juvenile Research, Department of Research, State of Illinois,

Howard, K. I. & Diesenhaus, H. I. Personality correlates of change-
seeking behavior. Perceptual and Motor Skills, 1965. 21, 655-664.

Johns, J. H. & Quay, H. The effect of social reward on verbal
conditioning in psychopathic and neurotic military offenders.
Journal of Consulting Psychology, 1962, 26, 217-220.

Jones, A. Stimulus-seeking behavior. In Zubek, J. (Ed.) Sensory
Deprivation: Fifteen Years of Research. Appleton-Century Crofts,
1969, 167-206.

Kingsley, L. Wechsler-Bellevue patterns of psychopaths. Journal of
Consulting Psychology, 1960, 24, 373.

Kish, G. B. & Busse, W. Correlates of stimulus-seeking: Age, education,
intelligence, and aptitudes. Journal of Consulting and Clinical
Psychology, 1968, 32, 633-637.

Lanyon, R. I. MMPI: A Handbook of MMPI Group Profiles. University of
Minnesota Press, Minneapolis, 1968.

Leuba, C. Toward some integration of learning theories: The concept
of optimal stimulation. Psychological Reports, 1955, 1, 27-33.

Luebeck, R. M. Sensory adaptation in psychopathic and neurotic delin-
quents. Unpublished doctoral dissertation, University of Iowa,

Lykken, D. T. A study of anxiety in the sociopathic personality.
Journal of Abnormal Psychology, 1957, 55, 6-10.

Mack, J. L. The MMPI and recidivism. Journal of Abnormal Psychology,
1969, 74, 612-614.

Maddi, S. R., Probst, B. & Feldinger, I. Three expressions of the need
for variety. Journal of Personality, 1965, 33, 82-98.

McCord, W. & McCord, Joan. The Psychopath: An Essay on the Criminal
Mind. Princeton: Van Nostrand, 1964.

Morello, M. A study of the adjustive behavior of prison inmates to
incarceration. Unpublished doctoral dissertation, Temple
University, 1959.

Myers, T. I. Tolerance for sensory and perceptual deprivation. In
Zube k, J. (Ed.) Sensory Deprivation: Fifteen Years of Research.
New York: Appleton-Century Crofts, 1969, 167-206.

Penney, R. K. & Reinehr, R. C. Development of a stimulus-variation
seeking scale for adults. Psychological Reports, 1966, 18, 631-638.

Persons, R. W. & Bruning, J. L. Instrumental learning with sociopaths:
A test of clinical theory. Journal of Abnormal Psychology, 1966,
71, 165-168.

Petrie, A. Individuality in Pain and Suffering. Chicago: University
of Chicago Press, 1967.

Quay, H. C. Psychopathic personality as pathological stimulation-
seeking. American Journal of Psychiatry, 1966, 122, 180-183.

Rankin, R. J. & Wikoff, R. L. The IES Arrow Dot performance of delin-
quents and nondelinquents. Perceptual and Motor Skills, 1964, 18,

Reinehr, R. C. The relation between scores on the stimulus variation
scale and autokinetic movement. Psychonomic Science, 1965, 3,

Rettig, S. Ethical risk sensitivity in male prisoners. British
Journal of Criminology, 1964, 4, 582-590.

Schacter, S. & Latane, B. Crime, cognition, and the autonomic nervous
system. In Jones, M. R. (Ed.) Nebraska Symposium on Motivation.
Lincoln: University of Nebraska Press, 1964.

Schuessler,K. F. & Cressey, D. B. Personality characteristics of
criminals. American Journal of Sociology, 1950, 55, 476-484.

Sherman, L. J. Retention in psychopathic, neurotic, and normal subjects.
Journal of Personality, 1957, 6, 721-729.

Sidle, A. Originality in problem solving as a function of anxiety and
withdrawal in schizophrenics. Journal of Consulting Psychology,
1963, 27, 550.

Skrzypek, G. J. Effect of perceptual isolation and arousal on anxiety,
complexity, preference, and novelty preference in psychopathic and
neurotic delinquents. Journal of Abnormal Psychology, 1969, 74,

Stennett, R. G. The relationship of performance level to level of
arousal. Journal of Experimental Psychology, 1957, 54, 54-61.

Suedfeld, P., Gluckaber, S. & Vernon, J. Sensory deprivation as a drive
operation: Effects upon problem solving. Journal of Experimental
Psychology, 1967, 75, 166-169.

Tranel, N. Effects of perceptual isolation on introverts and extroverts.
Journal of Psychiatric Research, 1962, 2, 185-192.

Waldo, G. P. & Dinitz, S. Personality attributes of the criminal: An
analysis of research studies, 1950-1965. Journal of Research in
Crime and Delinquency, 1967, 4, 185-202.

Watman, W. A. The relationship between "Acting out" behavior and some
psychological test indices in a prison population. Journal of
Clinical Psychology, 1966, 22, 3.

Weisen, A. E. Differential reinforcing effects of onset and offset of
stimulation on the operant behavior of normals, neurotics and psycho-
paths. Unpublished doctoral dissertation, University of Florida,

White, R. W. Motivation reconsidered: The Concept of competence.
Psychological Review, 1959, 66, 297-333.

van Wulfften Palthe, P. M. Fluctuations in level of consciousness
caused by reduced sensorial stimulation and by limited motility
in solitary confinement. Psychiatric neurologic Neurochrinug
(Amsterdam), 1962, 65, 377-402. In Zubek, J. Sensory Deprivation:
Fifteen Years of Research. New York: Appleton-Century Crofts,
1969, 265.

Zubek, J. Sensory and perceptual-motor processes. In Zubek, J. (Ed.)
Sensory Deprivation: Fifteen Years of Research. New York:
Appleton-Century Crofts, 1969a,207-253.

Zubek, J. Physiological and biochemical effects. In Zubek, J. (Ed.)
Sensory Deprivation: Fifteen Years of Research. New York:
Appleton-Century Crofts, 1969b, 254-288.

Zuckerman, M. Variables affecting deprivation results. In Zubek, J.
(Ed.) Sensory Deprivation: Fifteen Years of Research. New York:
Appleton-Century Crofts, 1969a, 47-84.

Zuckerman, M. A new form (IV) of the Sensation Seeking Scale. Unpub-
lished manuscript, University of Delaware, 1969.

Zuckerman, M., Lolin, E., Price, L. & Zoob, I. Development of a
sensation-seeking scale. Journal of Consulting Psychology, 1964,
28, 427-432.

Zuckerman, M. & Link, K. Construct validity for the Sensation Seeking
Scale. Journal of Consulting and Clinical Psychology, 1968, 32,

Zuckerman, M., Persky, H., Miller, L. & Levin, B. Controlling effects
of under-stimulation (Sensory Deprivation) and overstimulation
(High stimulus variety). Proceedings of the 77th Annual Convention,
American Psychological Association, 1969, Washington, D. C.


Appendix A

M. Zuckerman

DIRECTIONS: Each of the items below contains two choices, A and B.
Please indicate on your answer sheet which of the choices most describes
your likes or the way you feel. In some cases you may find items in
which you do not like either choice. In these cases mark the choice
you dislike least.

It is important that you respond to all items with only one choice,
A or B. We are interested only in your likes or feelings, not in how
others feel about these things or how one is supposed to feel. There
are no right or wrong answers as in other kinds of tests. Be frank
and give your honest appraisal of yourself.

1. A. I would like a job which would require a lot of traveling.
B. I would prefer a job in one location.

2. A. I am invigorated by a brisk, cold day.
B. I can't wait to get into the indoors on a cold day.

3. A. I find a certain pleasure in routine kinds of work.
B. Although it is sometimes necessary I usually dislike routine
kinds of work.

4. A. I often wish I could be a mountain climber.
B. I can't understand people who risk their necks climbing

5. A. I dislike all body odors.
B. I like some of the earthy body smells.

6. A. I get bored seeing the same old faces.
B. I like the comfortable familiarity of everyday friends.

7. A. I like to explore a strange city or section of town by myself,
even if it means getting lost.
B. I prefer a guide when I am in a place I don't know well.

8. A. I find the quickest and easiest route to a place and stick to it.
B. I sometimes 'ake different routes to a place I often go, just for
variety's sake.

9. A. I would not like to try any drug which might produce strange
and dangerous effect on me.
B. I would like to try some of the new drugs that produce

10. A. I would prefer living in an ideal society where everyone is
safe, secure and happy.
B. I would have preferred living in the unsettled days of our

11. A. I sometimes like to do things that are a little frightening.
B. A sensible person avoids activities that are dangerous.

12. A. I order the dishes with which I am familiar, so as to avoid
disappointment and unpleasantness.
B. I like to try new foods that I have never tasted before.

13. A. I can't stand riding with a person who likes to speed.
B. I sometimes like to drive very fast because I find it exciting.

14. A. If I were a salesman I would prefer a straight salary rather
than the risk of making little or nothing on a commission basis.
B. If I were a salesman I would prefer working on a commission if
I had a chance to make more money than I could on a salary.

15. A. I would like to take up the sport of water-skiing.
B. I would not like to take up water-skiing.

16. A. I don't like to argue with people whose beliefs are sharply
divergent from mine, since such arguments are never resolved.
B. I find people that disagree with my beliefs more stimulating
than people who agree with me.

17. A. When I go on a trip I like to plan my route and timetable
fairly carefully.
B. I would like to take off on a trip with no pre-planned or
definite route, or timetables.

18. A. I enjoy the thrills of watching car races.
B. I find car races unpleasant.

19. A. Most people spend entirely too much money on life insurance.
B. Life insurance is something that no man can afford to be without.

20. A. I would like to learn to fly an airplane.
B. I would not like to learn to fly an airplane.

21. A. I would not like to be hypnotized.
B. I would like to have the experience of being hypnotized.

22. A. The most important goal of life is to live it to the fullest
and experience as much of it as you can.
B. The most important goal of life is to find peace and happiness.

23. A. I would like to try parachute jumping.
B. I would never want to try jumping out of a plane, with or
without a parachute.

24. A. I enter cold water gradually giving myself time to get used
to it.
B. I like to dive or jump right into the ocean or a cold pool.

25. A. I do not like the irregularity and discord of most modern
B. I like to listen to new and unusual kinds of music.

26. A. I prefer friends who are excitingly unpredictable.
B. I prefer friends who are reliable and predictable.

27. A. When I go on a vacation I prefer the comfort of a good room
and bed.
B. When I go on a vacation I would prefer the change of camping

28. A. The essence of good art is in its clarity, symmetry of form
and harmony of colors.
B. I often find beauty in the "clashing" colors and irregular
forms of modern paintings.

29. A. The worst social sin is to be rude.
B. The worst social sin is to be a bore.

30. A. I look forward to a good night of rest after a long day.
B. I wish I didn't have to waste so much of a day sleeping.

31. A. I prefer people who are emotionally expressive even if they
are a bit unstable.
B. I prefer people who are calm and even-tempered.

32. A. A good painting should shock or jolt the senses.
B. A good painting should give one a feeling of peace and security.

33. A. When I feel discouraged I recover by relaxing and having some
soothing diversion.
B. When I feel discouraged I recover by going out and doing some-
thing new and exciting.

34. A. People who ride motorcycles must have some kind of an uncon-
scious need to hurt themselves.
B. I would like to drive or ride on a motorcycle.

Appendix B


Item Raw Score Analysis Percentage Analysis X2

Marion Menard SIU

1 B

2 A
2 B

3 A
3 B

4 A
4 B

5 A
5 B

6 A
6 B

7 A
7 B

8 A
8 B

9 A
9 B

10 A
10 B

11 A
11 B

12 A
12 B

13 A
13 B

Marion Menard SIU

73.7 67.5 52.5
26.3 32.5 47.5

100.0 50.0 55.7
0.0 50.0 44.3

21.1 25.0 21.3
78.9 75.0 78.7

63.2 55.0 63.9
36.8 45.0 36.1

26.3 20.0 31.1
73.7 80.0 68.9

63.2 25.0 14.8
36.8 75.0 85.2

89.5 87.5 80.3
10.5 12.5 19.7


32.5 27.9
67.5 72.1

68.4 70.0 68.9
31.6 30.0 31.1

36.8 50.0 59.0
63.2 50.0 41.0

78.9 82.5 78.7
21.1 17.5 21.3

15.8 22.5 42.6
84.2 77.5 57.4

15.8 40.0 36.1
84.2 60.0 63.9






< .1




<- 10


< 5


Appendix B (continued)

Item Raw Score Analysis Percentage Analysis X2

Marion Menard SIU

14 A
14 B

15 A
15 B

16 A
16 B

17 A
17 B

18 A
18 B

19 A
19 B

20 A
20 B

21 A
21 B

22 A
22 B

23 A
23 B

24 A
24 B

25 A
25 B

26 A
26 B

27 A
27 B

Marion Menard SIU

21.1 27.5 41.0
78.9 72.5 59.0

89.5 75.0 82.0
10.5 25.0 18.0

21.1 40.0 27.9
78.9 60.0 72.1

31.6 42.5 52.5
68.4 57.5 47.5

73.7 77.5 85.2
26.3 22.5 14.8

52.6 45.0 39.3
47.6 55.0 60.7

78.9 92.5 86.9
21.1 7.5 13.1

26.3 27.5 26.2
73.7 72.5 73.8

68.4 50.0 54.1
31.6 50.0 45.9

73.7 72.5 62.3
26.3 27.5 37.7

42.1 47.5 29.5
57.9 52.5 70.5

36.8 17.5 37.7
63.2 82.5 62.3

15.8 22.5 34.4
84.2 77.5 65.6

52.6 55.0 62.3
47.4 45.0 37.7

29 38'
11 23

4 10



< 10

4 10


< 10


.< 10


< 10

< 10

< 10


Appendix B (continued)

Item Raw Score Analysis Percentage Analysis X2

Marion Menard SIU

28 A
28 B

29 A
29 B

30 A
30 B

31 A
31 B

32 A
32 B

33 A
33 B

34 A
34 B

Marion Menard SIU

47.4 72.5 44.3
52.6 27.5 55.7

31.6 65.0 62.3
68.4 35.0 37.7

57.9 72.5 63.9
42.1 27.5 36.1

36.8 35.0 45.9
63.2 65.0 54.1

31.6 30.0 41.0
68.4 70.0 59.0

31.6 57.5 39.3
68.4 42.5 60.7

31.6 15.0 19.7
68.4 85.0 80.3

< 5

< 5




< 10

< 10

_ _

Appendix C



(a copy of this will be placed in your files)

Dear Participant,

Thank you for volunteering and taking part in the experiment. The
purpose of this study was to investigate some factors related to "Need
for Stimulation". It was based on the assumption that different indi-
viduals have varying degrees of need for stimulation from the environ-
ment and that this need is reflected in their behavior as well as their
personality. It was hypothesized that individuals who are more outgoing,
adventurous and risk-taking tend to have the greater stimulation needs.
We also investigated to see if criminality is related to sensation

You participated in one of two experimental conditions. (1)
Restriction of stimulation: spending hour blindfolded, with ear
plugs, while in a reclining chair or (2) Enhancement of stimulation:
spending hour watching an action movie. Before and after these condi-
tions, you were given three tasks. The Maze task involved your choosing
any of a number of paths all of which are correct. The more different
paths you chose, the higher your sensation seeking score. The circle
and line task is based on findings that sensation seekers tend to reduce
the size of stimuli over time. You scored high as a sensation seeker
if you underestimated the size of the line or circle as you progressed
from the first to the 4th trial. The Questionnaire had one sensation
seeking answer and one anti-sensation seeking answer for each question.

All of these results will be compared to the group answers obtained
from Southern Illinois University students to see if there is a differ-
ence in degree of sensation seeking between prisoners and other college

The results of this study will not be available for several months,
and of course, Individual scores will not be made available to anyone.
Thank you again for volunteering to participate in this study.


Annette M. Brodsky
Counseling and Testing Center
Southern Illinois University

Appendix D


May 24, 1969

Dear Student:

Thank you for participating in the experiment "Mazes and other
tasks". The purpose of the experiment was to study differences among
personality types in their need for sensory stimulation. It was based
on the finding that some people have a greater need for stimulation and
that this need is reflected in their personalities. They tend to be
outgoing, adventurous and risk-taking.

You participated in one of two experimental conditions...1. Restric-
tion of stimulation--spending hour blindfolded, with ear plugs in a
reclining position or 2. hour of stimulation--watching a movie in
the presence of others. The tasks you took before and after the experi-
mental condition have been demonstrated to be measures of need for
stimulation. It was given 2 times to also measure if change might
occur as a result of stimulation differences in the environment.

Part I of the experiment was the personality test used to differen-
tiate the personality types and patterns that might be related to need
for stimulation. This was accomplished by comparing personality scores
with high and low scorers on the stimulation measuring tasks.

Thank you for participating.


Annette M. Brodsky

Appendix E






Appendix F





Appendix G


Menard (N=40)

Bank Robbery
Rape and Assault
Assault w/Dang. Weap.
Att. Kidnapping
Assault & Att. to E3cape


Damage to Property
Aggressive Battery
Armed Robbery
Voluntary Manslaughter


convicted for more than one of these crimes.

Marion (N=19)


* Some Subjects were


Annette Mae Brodsky was born Annette Mae Ratner, January 23, 1938,

in Chicago, Illinois. In 1949 she moved to Coral Gables, Florida. She

attended Coral Gables High School and Miami Senior High School, where

she graduated in June, 1956. She received the degree Bachelor of Arts,

with a major in psychology, from the University of Miami in June, 1960.

She received the Master of Arts degree, in psychology, with a minor

in philosophy from the University of Florida, February, 1962.

She was commissioned as an MSC Officer in the United States Army,

June, 1963, and completed a clinical psychology internship at Walter

Reed General Hospital, Washington, D. C., August, 1964. She served

as staff psychologist at the U.S. Disciplinary Barracks, Ft. Leavenworth,

Kansas, from September, 1964 to May, 1965. Since September, 1967 she

has been employed as a counseling psychologist at the Counseling and

Testing Center, Southern Illinois University, Carbondale, Illinois.

She is married to Stanley L. Brodsky and is the mother of two


This dissertation was prepared under the direction of the chairman
of the candidate's supervisory coiattee end has been approved by all
members of that committee. It vas submitted to the Dean of the College
of Arts ead Sciencees nd to the Graduate Council, and was approved as
partial fulfillment of the requirements for the degree of Doctor of

March, 1970

Dean, College of Ae and S-ce

Degn, Graduate School

Supervisory Committee:


^ ^___..