Offender variables and treatment outcomes of participants in a residential sex offender program


Material Information

Offender variables and treatment outcomes of participants in a residential sex offender program
Physical Description:
vii, 146 leaves : ; 29 cm.
Shaw, Theodore A., 1945-
Publication Date:


Subjects / Keywords:
Sex offenders -- Rehabilitation -- Florida   ( lcsh )
Psychotherapy   ( lcsh )
Behavior therapy   ( lcsh )
Counselor Education thesis Ph. D
Dissertations, Academic -- Counselor Education -- UF
bibliography   ( marcgt )
non-fiction   ( marcgt )


Thesis (Ph. D.)--University of Florida, 1991.
Includes bibliographical references (leaves 138-145).
Statement of Responsibility:
by Theodore A. Shaw.
General Note:
General Note:

Record Information

Source Institution:
University of Florida
Rights Management:
All applicable rights reserved by the source institution and holding location.
Resource Identifier:
aleph - 001722860
oclc - 25769045
notis - AJD5362
sobekcm - AA00004754_00001
System ID:

Full Text







Copyright 1991


Theodore A. Shaw


I would like to acknowledge the following

individuals: my doctoral chairperson, Dr. Ellen

Amatea, for her never ending patience and perseverance;

my doctoral committee members including Dr. George

Barnard, Dr. Harry Grater, Dr. Larry Loesch, and Dr.

Gus Newman for their support and guidance; Lynn Robbins

for her friendship and invaluable assistance; Denny

Gies and the administration and clinical staff of North

Florida Evaluation and Treatment Center; and the

dedicated staff of the Sex Offender Unit of NFETC.

I would also like to acknowledge my wife, Dr.

Jamie Funderburk, for her patience and continuing

support, guidance and technical assistance, my

children, Amanda and Alex, for their patience with my

preoccupation, and my mother, Micki, who has always

inspired me to greater achievement.




ACKNOWLEDGMENTS............................. ......... iii


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

Scope of the Problem........................ 4
Need for the Study......................... 13
Purpose of the Study...................... 14
Research Questions......................... 15
Context for the Study..................... 16
Definition of Terms....................... 18
Significance of the Study.................. 20

2 REVIEW OF LITERATURE...................... 22

Research on Sex Offender Treatment
Amenability.............................. 22
Basic Assumptions, Techniques, Settings,
Populations Treated (Focus), Research
Findings and Related Literature......... 36
Treatment Center Study/Setting............. 45
Treatment Amenability Factors............. 57
Independent Variables..................... 62
Treatment Outcome Categories.............. 72
Summary..................................... 75

3 METHODOLOGY................................ 76

Research Design ........................... 77
Population ............................... 80
Resultant Sample.......................... 82
Procedures................................ 83
Instrumentation......................... 91
Hypotheses................................. 96

Data Processing and Analysis.............. 97

4 RESULTS .................................. 98

Descriptive Statistics for Treatment
Outcome Groups........................... 99
Differences Among Treatment Outcome
Groups.................................... 100
Summary ................................... 114

5 DISCUSSION................................ 116

Discussion of Results..................... 116
Limitations............................... 123
Implications.............................. 127
Summary and Conclusions................... 135

APPENDIX.......................................... 136

REFERENCES........................................ 137

BIOGRAPHICAL SKETCH.. ............................. 144

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



Theodore A. Shaw

December, 1991

Chairperson: Ellen Amatea, Ph.D
Major Department: Counselor Education

Growing awareness of the frequency of sexual abuse

and its terrible impact on victims and society has led

to the development of more systematic approaches to the

evaluation and treatment of sex offenders. Little is

known concerning which treatment models and settings

are effective with which populations of sex offenders.

This study was conducted to examine the relationships

among six offender variables and offender outcome

status in a residential sex offender treatment program

which emphasized a multi-modal treatment approach. The

six offender variables were (a) age, (b) reading

ability, (c) degree of discrimination in the selection


of victim and offense type, (d) instant offense

(offense for which the offender is currently

incarcerated), (e) presence of antisocial personality

disorder, and (f) marital status. The three

categories of treatment outcome were (a) not amenable

for treatment, (b) amenable for treatment but

discharged with a poor prognosis, and (c) amenable for

treatment and discharged with a good prognosis.

Results of a discriminant function analysis revealed

that the six offender variables together did not

discriminate significantly among offenders in the three

outcome groups. Results of analyses of variance and

Chi-square analyses revealed that offenders who were

married or in a committed relationship and offenders

who had higher reading ability were more likely to be

found in the treatment group discharged with a good

prognosis than in the other two treatment status

groups. Presence of antisocial personality disorder,

degree of discrimination in offense or victim type,

age, and instant offense did not discriminate

significantly among offenders in the various outcome

groups. Limitations, implications for theory and

clinical practice, and recommendations for future

research are also discussed.



The challenge of treating sex offenders has taken

on new importance in light of the staggering reports on

frequencies of types of sexual abuse and rising

awareness of the human and financial costs of coping

with the problems ineffectively. Of particular concern

are the increased reports of rapes of adult women and

men as well as reports of intrafamilial and

extrafamilial child sexual abuse. Recent research

reveals that a pattern of repeated, unreported offenses

is more the rule than the exception among sex offenders

(Barnard, Fuller, Robbins, & Shaw, 1989; Finkelhor,

1988; Furby, Weinrott, & Blackshaw, 1989; Knopp, 1984;

Salter, 1988). In their study of 411 parapheliacs,

Abel and associates discovered that these subjects had

engaged in over 138,000 sex offenses involving over

115,000 victims (Abel, Becker, Mittelman, Cunningham-

Rathner, Rouleau, & Murphy, 1987). Moreover, many had

never been arrested, few had ever been incarcerated,

and most had never received specific sex

offender treatment.

The enormous financial and psychological costs to

victims and society are major factors that underscore

the need to treat sex offenders. Through reports by

sociological researchers such as Finkelhor and

clinicians working in the relatively new Adults

Molested As Children (AMACS) movement, it is clear that

vast numbers of men and women have been sexually

victimized and that such victimization has profound

traumatic effect on these individuals (Burgess, Groth,

Holmstrom, & Sgroi, 1978).

Growing awareness of sex offense problems has led

to the development of many specialized treatment

program efforts for sex offenders. For example, a

national survey of adult and juvenile residential and

outpatient sex offender programs conducted by Knopp and

Stevenson in 1989 revealed that between 1986 and 1988

the number of specialized treatment programs increased

by 56% nationwide, to a total of more than 1000.

Relatedly, of the adult treatment services polled, 28%

(119) were residential, the majority of which were

located in adult prisons, and 72% (310) were

outpatient, the majority of which were provided by

mental health private practitioners. There was a

similar distribution for juvenile services: 20% (117)

were residential, with almost half located in private

facilities, and 80% (456) were outpatient, with about

the same number provided by private practitioners and

mental health centers. However, the demand for

specialized sex offender treatment continues to outpace

its availability. In Florida, for example, by 1989

more than 5000 sex offenders were incarcerated in the

prison system. Of these, more than 300 were found

eligible (i.e., met entry criteria) for treatment and

had been placed on a waiting list. Although only 75

offenders were admitted to treatment that year, more

than 150 had been placed on the waiting list. As in

Florida, many states are facing situations where

treatment opportunities are limited by funding which

determines the number of available treatment beds.

With many more offenders in need of treatment than

available beds, it becomes crucial to determine which

offenders might profit most from specialized treatment.

There is little empirical evidence available,

however, concerning who is most likely to benefit from

sex offender treatment (Abel, Mittelman, Becker,

Rathner, & Rouleau, 1988; Laws, 1985), and few studies

have been undertaken to determine treatment specific

amenability. Because most of the current research on

sex offender treatment has focused only on examining

official and unofficial reports of rearrest,

reconviction, or self-report of reoffense as measures

of treatment outcome, the present state of knowledge

regarding amenability in terms of interim measures such

as the offender's status at the time of discharge from

treatment is relatively sparse. Therefore, the

correlation among six offender characteristics and

three sex offender treatment outcomes was examined in

this study.

Scope of the Problem

Because the development and widespread use of

specialized treatment procedures for sex offenders are

relatively recent, very little is known regarding

either the types of treatment procedures most effective

with particular types of sex offenders or

characteristics which distinguish offenders who benefit

from particular treatments from those who do not.

Although syndromes relating to sexual deviance have

been known to the medical/psychological field for

centuries, organized approaches to treating and

assessing the impact of treatment of this population

are relatively recent (Barnard et al., 1989; Greer &

Stuart, 1983).

The problem with outcome research in the sex

offender treatment area is less one of quantity than

methodology, particularly in regard to outcome

measures. In a recent review of the state of tne art

of sex offender treatment outcome research, Furby et

al. (1989) concluded that it has not been adequately

demonstrated that specialized sex offender treatments

reduce recidivism more effectively than other

strategies (including no treatment).

Recently, researchers have begun to evaluate

critically both the quantity and quality of research on

sex offender treatment. For example, in reviewing the

recent research on sex offender treatment, Murphy

(1988) contended that current providers of sex offender

treatment still know very little about the efficacy of

any of their treatment approaches and that much more

research needs to be done to speak with assured

confidence about the nature of specialized treatment of

sexual abusers. In commenting on the lack of

definitive research in this area, Murphy (1988)

criticized the methodology used in sex offender

treatment research and proposed a set of questions to

guide future research. Among these was, "Which

offender characteristics predict amenability for which

treatment?" Other researchers have concurred that a

key issue in improving sex offender treatment efficacy,

as well as in development of clearer understanding of

treatment impact for sex offenders, centers on

identifying those individuals most likely to benefit

from treatment (Abel et al., 1989; Furby et al., 1989;

Knopp, 1984; Marshall & Barbaree, 1988, 1990; Murphy,

1988; Shaw, Spears, Cunningham, Butler, Barnard,

Robbins, & Newman, 1987).

Only a limited number of researchers have examined

the characteristics of offenders as a means of

predicting success in completing sex offender

treatment. Further, most researchers who have examined

offender characteristics have focused predominantly on

outpatient populations (in the community) and have done

so secondarily to some other research focus. Their

findings from outpatient samples suggest that offenders

who are relatively more intelligent (Marshall &

Barbaree, 1988), have stable marital relationships

(Abel et al., 1988), have committed only one type of

sex offense on only one type of victim (e.g., among

pedophiles only) are likely to complete treatment

successfully and have relatively low recidivism rates

(Abel et al., 1988). Little is known, however,

regarding the amenability for treatment of incarcerated

offenders in secure treatment programs (i.e., prisons

or maximum security mental institutions) where fear of

reprisals, presumed relatively high prior offense

rates, low socioeconomic status, and minimal

educational attainment are likely to be factors

negatively influencing amenability (Knopp, 1984).

Populations of incarcerated offenders are likely to

contain fixated (i.e., repetitive) pedophiles, rapists,

and other violent offenders who are routinely excluded

from outpatient treatment due both to the obvious risk

to the community and their high likelihood, even on a

first conviction, to be incarcerated instead of

receiving probation or community control based on the

perceived seriousness of their offenses (Knopp, 1984).

Because of these factors, the findings of studies based

on outpatient samples appear to have limited

generalizability to incarcerated offenders.

Despite limited empirical investigation, however,

it has been customary for practitioners treating

incarcerated sex offenders to assume that not all sex

offenders are amenable for treatment and to screen

potential treatment candidates based on these

assumptions (Schwartz, 1988b). Because there are

limited empirical data, guidelines for determining

amenability for treatment usually have been established

on the basis of clinical judgment, tradition,

limitations of the treatment setting (including

therapeutic skills of the clinicians), or funding

considerations. New Jersey, Minnesota, and Oregon, for

example, are among the few states that offer rapists

the opportunity for treatment because they have long-

established, secure residential treatment facilities

(Knopp, 1984).

Among the assumptions commonly used to select

offenders for treatment is the notion that sex

offenders with antisocial personality disorders are

less likely to complete treatment successfully due to

their documented history of frequent job changes,

adolescent conduct problems and unwillingness to follow

rules (Abel et al., 1988).

Sex offenders have been categorized traditionally

according to their most recent charge (instant offense)

or most common offense if the instant offense is

uncharacteristic of the documented offense history.

Incest offenders are generally considered the most

amenable for treatment as well as the least likely to

reoffend whether they receive treatment or not. Child

molesters are considered less amenable for treatment

than incest offenders but more amenable than rapists.

Rapists are generally excluded from treatment because

they frequently have a diagnosis of antisocial

personality disorder and because of the nature of their

offenses (which makes them "too dangerous" for

outpatient settings) (Knopp, 1984; Schwartz, 1988b).

Moreover, rapists are often physically aggressive and

threatening and are likely to engage in rule violations

which could cause them to be prematurely terminated

from treatment. A widespread clinical assumption in

the field is that as offenders get older they become

less likely to engage in aggressive behaviors and more

concerned with the consequences of their criminal

behavior. It follows that the older the offender, the

more likely he is to make the necessary effort to

complete treatment successfully.

Although not demonstrated definitively in the

literature, intelligence (IQ) and reading level also

have been assumed to be important predictors of

amenability for sex offender treatment (Marshall &

Barbaree, 1988; Rosen, 1964; Schwartz, 1988b). The

fact that there are still virtually no treatment

programs for retarded sex offenders is one indication

that a certain level of mental ability is considered

important in the treatment process. As

cognitive/behavioral strategies have become more

integral to sex offender treatment, adequate reading

ability and intelligence have become important criteria

for program admission. It has been assumed that the

more intelligent the treatment candidate and the better

able to read, the more likely the individual is to

complete treatment successfully. Requirements for

successful completion of many treatment programs

include compliance with extensive reading and related

homework assignments and integration of the learned

material into daily behavior patterns (North Florida

Evaluation and Treatment Center, 1988; Rosen, 1964).

A fifth assumption undergirding the selection of

offenders for treatment concerns the sex offender's

degree of discrimination in offense and victim types

and its impact on treatment participation and success.

In a recently published study Abel and associates

(1988) found that lack of discrimination in offense and

victim types predicted dropout of offenders from an

outpatient treatment program. In addition, this

variable predicted recidivism for those who did

complete treatment. Significantly, more than half of

the participants in the study of Abel et al. (1988)

reported multiple deviations. This assumption has been

considered only recently by clinicians and is supported

by empirical evidence (Marshall & Barbaree, 1988).

Suggesting that the disorder of individuals with

multiple victim and offense types may be more

entrenched in their deviant behaviors and thus

resistant to change through treatment has important

implications for determining amenability for treatment.

This finding from a community-based program population

has potential significance for residential programs

which are likely to be treating those with documented

reoffenses and long histories of deviant behavior

(Knopp, 1984).

A sixth assumption regarding selection of

treatment candidates which has recently received

research support is marital status. It has long been

thought by clinicians that offenders with intact,

supportive families were more likely to both

successfully complete treatment and refrain from

reoffending, and this may contribute to the incest

offender's reputation for successful treatment outcome.


Recently Abel and associates found that marital status

was the single demographic variable which predicted

both dropout and recidivism. Those who were married

were more likely to complete treatment and less likely

to reoffend than offenders who were separated, divorced

or never married (Abel et al., 1988).

In summary, despite the absence of empirical

corroboration, practitioners in the field of sex

offender treatment have routinely based their selection

decisions on a number of shared assumptions. These

assumptions are that persons who are single or divorced

and who have low intelligence, poor reading ability, an

antisocial personality disorder, history of violent

sexual assaults (i.e., rape), a documented repetitive

history of offending, and history of different types of

sex offenses against victims of differing age or sex

will be less likely to complete treatment successfully.

Conversely, most practitioners have assumed that those

who are married, have high intelligence, adequate

reading ability, lack of antisocial personality traits,

absence of a history of violent sexual assaults or

repetitive offending, and who engage in a single type

of offense (i.e., incest) against a particular age

group of a particular sex (e.g. pre-pubertal boys) will

be more likely to successfully participate in


Need for the Study

In virtually every state in America the number of

sex offenders in need of treatment far exceeds the

available treatment opportunities (Knopp, 1984; Salter,

1988). While more and more treatment programs are

opening in an effort to meet this need, there exist no

empirically derived selection standards for these

programs. Nonetheless, program clinicians and

administrators are routinely expected to determine

amenability and provide a treatment regimen. Although

numerous researchers and clinicians have offered

reasonable models for determining amenability (Barnard

et al., 1989; Salter, 1988; Schwartz, 1988b), at the

present time these models are untested. Such a lack of

research evidence on amenability has a distinctly

negative impact on the ability of programs to maximize

their efforts and to replicate their successes or

improve their treatment programs. Thus, more needs to

be learned concerning what types of treatment are most

effective with which types of offenders (Schwartz,

1988b). Studies must be undertaken in a variety of

settings to test the generalizability of the findings

reported from the few studies which have been

conducted. Although several researchers have recently

presented data on amenability for treatment, these

programs have been community based and limited to

incest offenders or pedophiles (Abel et al., 1988;

Dwyer & Ambersson, 1985). Little has been demonstrated

concerning indicators of amenability for treatment of

populations of incarcerated sex offenders.

As has already been noted, leaders in the sex

offender treatment field concur that more research

needs to be undertaken in an effort to predict

successfully who will benefit from the emerging

specialized treatment models.

Purpose of the Study

This study was designed to assess the degree of

correlation among six offender characteristics and

three treatment outcome groups in a residential sex

offender treatment program. The six offender

characteristics were (a) reading level, (b) presence of

an antisocial personality disorder, (c) degree of

discrimination in offense and victim types

(operationalized as "total deviance score"), (d) age,

(e) marital status, and (f) instant offense (offense

for which the offender is currently incarcerated). The

three outcome groups consisted of (a) those not

accepted after an 8-week pretreatment evaluation phase,

(b) those accepted for treatment and subsequently

discharged from the treatment program with a poor

prognosis, and (c) those accepted for treatment and

subsequently discharged from treatment with a good


Research Questions

The following set of research questions were

addressed in this study:

1. What is the influence of reading level,

degree of discrimination in the choice of victim or

paraphilic act, presence of antisocial personality

disorder, age, marital status, and instant offense in

classifying offenders into outcome groups?

2. Are there differences in reading level of the

offenders among the three outcome groups?

3. Are there differences in the degree of

discrimination in the choice of offense and victim type

among the three outcome groups?

4. Are there differences

antisocial personality disorder


5. Are there differences

outcome groups?

6. Are there differences

the three outcome groups?

7. Are there differences

among the three outcome groups?

in the presence of an

among the three outcome

in ages among the three

in marital status among

in instant offense

Context for the Study

The setting in which the study was conducted is a

maximum security residential mental health treatment

center operated by the State of Florida Department of

Health and Rehabilitation Services (HRS) which offered

a model, structured sex offender treatment program. In

the same institution there were programs designed to

return offenders found Incompetent to Proceed (ITP) to

court and to treat until deemed no longer dangerous

offenders found by the court Not Guilty by Reason of

Insanity (NGI); those successfully treated might be

discharged to the community or some other less

restrictive setting.



The sex offender program had recently been revised

(Barnard et al., 1989) and has been described as a

model for the multi-modal or integrated approach to sex

offender treatment in a residential setting (Knopp,

personal correspondence, 1988). Although this implies

the use of many methods in an integrated whole, the

predominant treatment theories underlying the program

were cognitive/behavioral and psychodynamic.

Cognitive/behavioral sex offender treatment programs

are generally characterized by structured learning

modules built around a Relapse Prevention foundation

(Barnard et al., 1989). This type of treatment relies

heavily on the offender's ability to read, learn, and

understand reading assignments as well as his ability

to integrate group experiences and demonstrate an

integration of the covered material. Its location in a

maximum security forensic mental health facility with

twenty-four hour staff supervision and extensive

electronic controls made it hard for offenders not

interested in fulfilling the spirit of treatment and

integrating treatment material to make reasonable

progress. For this reason it is assumed that it was

unlikely that repetitive offenders or those with an

extensive criminal history and likely concommitant

antisocial personality disorders would have been able

to successfully complete the program.

Definition of Terms

Multi-Disciplinary Treatment Team (MDT) acceptance

staffing is a formal meeting with the offender to

review his general behavior within the milieu,

compliance with treatment plan directives, knowledge of

material covered in the evaluation phase modules,

results of any psychological testing, and any other

data relevant to the demonstration of his amenability

for treatment. The staffing occurred at the conclusion

of an eight week evaluation phase, or during the phase

at the request of the offender or his primary


Multi-Disciplinary Treatment Team (MDT)

termination staffing is a formal meeting of the MDT to

review the offender's progress in treatment, status at

the time of termination from treatment, and

recommendations after discharge.

Amenability for treatment is the likelihood of an

individual to benefit from a treatment procedure in

some describable or measurable way.

Evaluation phase refers to the first 8 weeks of

treatment in the Mentally Disordered Sex Offender

(MDSO) Program at North Florida Evaluation and

Treatment Center (NFETC), Gainesville, Florida.

Instant offense is the most recent offense or

group of offenses for which the offender has been

convicted and sentenced.

Recidivism has been variously defined as a

reoffense, rearrest, reconviction or reincarceration.

When not otherwise specified in the text, recidivism is

herein defined as a reoffense. Recidivism is

frequently used as a measure of treatment success with

criminals and individuals with repetitive or addictive

behaviors including deviant sexual behavior.

Reading level as measured by the Wide Range

Achievement Test (WRAT-R), can be described by a raw

score or a grade equivalent.

Age refers to how old the offender was upon

admission to NFETC.

Marital status refers to whether the offender was

married, divorced, separated or never married at the

time of admission to NFETC.

Antisocial personality disorder is a diagnostic

category from the personality disorder section of the

Diagnostic and Statistical Manual III (DSM III) (now

the Diagnostic and Statistical Manual III Revised), a

standard guide in the field of psychology and

psychiatry for making diagnoses of patients, which is

characterized by a history of delinquent behavior prior

to age eighteen followed by a history of irresponsible

behavior which might include criminal behavior,

substance abuse, frequent job changes, etc.

Degree of discrimination in offense and victim

types (total deviance score) refers to how many

different types of victims (e.g., pre-pubescent males

vs. pre-pubescent females) the offender has had and how

many different types of offense types (e.g., hands-on

vs. hands-off) the offender has engaged in.

Significance of the Study

The long-term consequences of systematically

studying the characteristics of those who do not appear

to profit from sex offender treatment are varied.

Learning who is more likely to benefit from a

particular treatment model has some very clear value,

particularly in a program that serves as a model for

the latest evolution in residential sex offender

treatment (Barnard et al., 1989). The cost of sex

offender treatment is significantly higher than the

cost of incarceration but, if effective in reducing

recidivism, treatment could prove to save society both

significant financial burdens and severe emotional and

physical trauma associated with virtually every

reoffense (Prentky & Burgess, 1990). Not only is sex

offender treatment more costly than incarceration, but

it is not available to all sex offenders. In fact,

even in systems with an extensive screening process and

a policy requiring volunteers, there are waiting lists

and offenders who request but never receive treatment.

This makes all the more important systematically and

correctly identifying those offenders who are most

likely to benefit from treatment.

Given the diversity of treatment programs

throughout the world and the lack as yet of a

recognizable standard for either screening or

treatment, it is essential that studies such as this

one be undertaken to attempt to determine who is most

likely to benefit from a specific treatment.


This chapter describes the research and practice

literature concerning sex offender treatment

amenability. Literature describing the characteristics

of incarcerated sex offenders is reviewed. In

addition, assumptions commonly held by practitioners

regarding amenability of sex offenders for treatment

will be discussed. Finally, the literature regarding

the predictor variables examined in this study is


Research on Sex
Offender Treatment Amenability

With society's growing awareness of the terrible

effects of sexual abuse have come greater demands on

the mental health treatment, correctional, and judicial

fields to deal effectively with sexual offenders. One

response to these demands has been the significant

increase in treatment programs available for offenders

(Knopp, 1984; Knopp, Rosenberg, & Stevenson, 1986;

Knopp & Stevenson, 1989). A second response has been


the development of national and international networks

of professionals to share treatment ideas (e.g.,

Association for the Treatment of Sexual Abusers [ATSA]

and National Adolescent Perpetrator Network [NAPN]).

Unfortunately, there is little empirical evidence

available regarding the level of effectiveness of sex

offender treatment programs or the appropriateness of

selection procedures (Earls & Quinsey, 1985; Furby et

al., 1989). With few exceptions, professionals

implementing sex offender treatment programs do not

evaluate their programs systematically but rather do

what they consider correct, often basing their

decisions as to who gets treatment and how they are to

be treated more on clinical intuition and assumptions

than on empirical evidence. Ultimately, the field of

sex offender treatment must answer its own specificity

question: What types of offenders, in what types of

treatment contexts, with what types of treatment

experiences/components, demonstrate what levels of

"cure" both initially and ultimately? This question

has been extremely difficult to answer with regard to

psychotherapy in general (Finkelhor, 1988; Knopp,

1984). Thus it is not surprising that there are so few

answers regarding either the appropriateness or the

effectiveness of treatment of sex offenders.

Although there is little empirical evidence

regarding treatment amenability based on sex offender

recidivism, there is even less empirical evidence

regarding amenability for treatment of offenders based

on treatment program outcome (Furby et al., 1989;

Marshall & Barbaree, 1990). Recognizing that

ultimately it will be necessary to consider a

comprehensive set of factors in answering the

specificity question for sex offender treatment, there

is a need for modest studies which can provide a basis

for more ambitious studies in this area (Earls &

Quinsey, 1985; Furby et al., 1989). Murphy (1988) has

proposed a set of questions that target variables which

partialize the specificity question for sex offender

treatment and thus can serve as a guide to research.

Among these is the question of who is amenable for what

types of treatment. This interim measure, which has

been alluded to in several papers and studies (Abel et

al., 1988; Green, 1988), is important for a number of

reasons. First, given the universally limited sex

offender specific treatment resources and their

frequently high cost, there is a pressing need to use

these resources efficiently. In 1989 in Florida, more

than 300 offenders were routinely maintained on a

waiting list; many of these offenders were discharged

from incarceration without treatment as a result of the

long waiting list (Alcohol, Drug Abuse and Mental

Health Program Office, Florida Department of Health and

Rehabilitative Services, 1989). Knowing who is most

likely to benefit from treatment or conversely who is

most likely to fail early on or drop out during the

course of treatment could maximize the treatment

prospects for offenders likely to benefit.

Although clinical assumptions abound, little

empirical evidence is available to provide direction

for developing sound amenability criteria for treatment

programs. Concommitantly, there is a recognized

absence of reliable outcome data for the variety of

treatment methods and program models in operation today

(Furby et al., 1989; Knopp, 1984). Earls and Quinsey

(1985), noted Canadian researchers and clinicians, in

summing up their conclusions regarding the need for

effective assessment and treatment of sex offenders,

suggest that there is a clear need for modest studies

to lay the foundation necessary to answer the pressing

questions in the sex offender treatment field. Among

other issues Earls and Quinsey (1985) suggest that the

specificity question has yet to be addressed in earnest

in the sex offender treatment research literature.

Short-term treatment outcomes are frequently

described as program evaluations; bureaucratic agencies

who routinely engage in program evaluations and

"quality assurance" reviews tend to focus on areas

other than treatment outcome except in terms of

percentages of admissions. For example, the 19d0

Report to the Florida Legislature concerning the two

residential sex offender programs focuses on a

comparison of the percent of graduates compared to the

total admissions. The assumption is that a "good"

treatment program would have a high percentage of

graduates. Yet experts stress the need for short-term

outcome measures in order to develop accurate measures

of treatment efficacy and specificity. Dr. Green


Another important reason for making program
evaluation an integral part of sex offender
treatment programs centers around the fact that
there actually exists very little scientifically
collected information regarding the efficacy of
sex offender treatment, in terms of recidivism and
behavior change. At this time, it has not been
established which treatment works best with what
type of offender, fundamental data which is
potentially vital to program managers, agency
policy makers, and legislators. (Green, 1988,

There are a number of factors which may, in part,

account for the dearth of research on sex offender


treatment selection. Perhaps most significantly, there

is little consistency in the sex offender treatment

field regarding criteria for determining amenability

for treatment, amenability being a combination of

initial eligibility for inclusion in a given treatment

program and the ability to progress satisfactorily and

attain some specified outcome, usually treatment

program specific. A number of factors have contributed

to this absence of consistency. First, the haphazard

development of treatment programs in isolated settings

around the United States and Canada has led to a

diversity of treatment models and methods and an array

of amenability criteria which are only program specific

(Knopp, 1984). Since there is no single model for the

treatment of sex offenders which is accepted by all

clinicians and there is reason to assume that

amenability may be program specific, it follows that

there exists no accepted method of determining

eligibility for treatment (Knopp, 1984). A second

reason is that the consistency which does exist

regarding salient criteria appears to emanate more

often than not from clinicians' assumptions rather than

empirical evidence (Knopp, 1984).

Still another factor which complicates the issue

of amenability for sex offender specific treatment is


the variety of external conditions with which programs

must contend. Constraints imposed by outside systems

within which the program functions may include (a)

specific populations from which treatment candidates

are selected (such as prisons, courts or other

institutions); (b) specific directives from funding

sources (such as targeting only the most "disturbed"

offenders still eligible for treatment); (c) specific

treatment models (which may be more difficult for some

individuals to complete); (d) particular settings in

which the program is provided including outpatient,

nonsecure residential, secure residential and prison;

(e) length of treatment (which may disqualify certain

offenders because they do not have enough time as in

the case of treatment programs for incarcerated

offenders where some offenders [e.g., youthful

offenders] routinely receive very short periods of

incarceration during which there is insufficient time

to be screened for treatment, transferred to a program

and complete it); and (f) competing philosophical


There is little question that the specific

treatment model/approach employed by the treatment

program is likely to have an effect on who is

considered amenable for that treatment.

Most programs recognize the limitations of the
clinical methods currently used with sex
offenders. Thus, they impose eligibility criteria
which reflect the perceived strengths and
limitations of their treatment programs. (Smith,
1988, p. 40)

For example, several authors have noted that a

particular model may require some minimum reading

skills or the ability to integrate material into daily

life (Barnard et al., 1989). Impacting on this dilemma

are the competing theories regarding treatment

candidate selection which differentially affect

eligibility criteria and which can be found throughout

the literature in the sex offender field. One theory,

championed by A. Nicholas Groth, argues that the most

dangerous and/or most likely to reoffend candidates

should be treated to the exclusion, if necessary, of

more amenable, less dangerous candidates (Knopp, 1984).

Programs, generally in secure, residential settings,

which have followed this strategy tend to have

relatively high rates of reoffense, if any, and have

had difficulty defending their position when funding

considerations arise (North Florida Evaluation and

Treatment Center, 1988a) because this model tends to

emphasize dangerousness/liklihood of reoffense over

amenability. A competing theory favors a triage model

which specifies that some criteria should be used to

determine who is most likely to benefit from treatment

and who is less likely to successfully complete or

significantly benefit from treatment. An example of

the latter is an outpatient treatment model which may

exclude individuals with extensive histories of

violence or previous crimes (Barnard et al., 1989).

A fourth contributing factor is the lack of

financial support for the testing of assumptions

regarding who is amenable for treatment. Because sex

offender treatment is not popular, most treatment

programs struggle to function with minimal and

unpredictable funds. Thus, although many new programs

are designed with a research component, funding cuts

often result in the elimination of such research

efforts. Therefore, it is not uncommon for programs to

have no organized approach to receiving feedback on

treatment efficacy and to focus only on publicized or

special cases of reoffending on the one hand and

standard case management measures such as progress

notes, treatment plans, and daily behavior while in

treatment on the other. Furthermore, the cost of

acquiring re-arrest data (a measure of treatment

efficacy) in both staff time and actual money and the

need to report reoffenses if they are revealed by

offenders or their families further block efforts at

systematic program evaluation.

A fifth contributing factor is that although

several studies have been published attempting to

describe, delimit or delineate one or another subtype

of sex offender (Stermac, Segal, & Gillis, 1990), the

findings of such studies reveal that sex offenders are

surprisingly heterogeneous. It appears they are linked

primarily by their offense and little else. For

example, when demographic variables such as income,

education, marital status, age, and intelligence were

examined, several studies have found results which

follow normal distributions (Abel et al., 1988; Shaw,

Barnard, Robbins, Spears, Cunningham, Butler, & Newman,

1988); incarcerated sex offenders differ little on

these same variables from other populations of

incarcerated offenders (Stermac et al., 1990). This

heterogeneity becomes yet another factor inhibiting the

delineation of specific eligibility criteria as this

population is difficult to differentiate from both

other criminal populations as well as among its own


Consequently there is very little research

literature available on treatment amenability in the


sex offender treatment field. That which is available

is limited to pedophiles treated in outpatient programs

(Furby et al., 1989). Given this limitation, there is

no assurance that any set of eligibility or amenability

criteria will be effective in identifying the best

candidates for a particular treatment program (zurby et

al., 1989).

Recently, Abel et al. (1988), reporting data

consisting of a one year follow-up using self-report of

reoffense, the dependent variable, as a measure of

recidivism from a sample of pedophiles being treated in

the community, found only one demographic variable to

be predictive of reoffense: marital status.

Demographic variables which did not distinguish between

nonrecidivists and recidivists were age, race, social

class, education, employment status, religious

preference, motivation for seeking treatment, frequency

of pedophilic acts before entering treatment, lifetime

number of molestation or sexual victims, and reported

self-control over pedophilic behavior before entering


When the multiplicity of age categories, gender
categories and hands on versus hands off
categories of pedophilic behavior were combined,
the combination proved to be a very significant
predictor of recidivism, with p < .0001. (Abel et
al., 1988, p. 230)


Although a significant, ongoing study, the limitations

of treating exclusively pedophiles in a community

setting makes these data unrepresentative of

incarcerated treatment populations which often include

as many as one third rapists and where as many as half

the offenders have an antisocial personality disorder

diagnosis (Barnard et al., 1989). Moreover, the

finding of marital status as a predictor of recidivism,

while it must be tested with incarcerated offenders,

bodes ill for incarcerated populations, who, for the

most part, are single, separated or divorced (Barnard

et al., 1989).

Treatment outcome data have been reported by

several other researchers with outpatient populations

(Abel et al., 1988; Maletsky, 1987; Marshall &

Barbaree, 1988; Knopp, 1984). Marshall and Barbaree

(1990) found that although higher intelligence

predicted successful outcome, all programs reported

relatively low recidivism rates (under 15%).

Randy Green, clinical'director in the sex offender

treatment field from the Forensic Services Unit of

Oregon State Hospital stated:

The only other serious experimental controlled
study in the field of sex offender treatment in
this country is being conducted in Tampa, Florida

by Richard Laws. Sponsored by the National
Institute of Mental Health (NIMH), Laws began this
outpatient study in August 1986, comparing effects
of a combined regimin of Relapse Prevention and
aversive conditioning, in contrast to a
traditional therapeutic model. (1988, p. 64)

Unfortunately, this study was prematurely terminated

primarily due to inability to achieve the necessary

sample size within the specified time, and federal

funds were withdrawn (D. R. Laws, personal

communication, November, 1989).

Furby's review of empirical studies of sex

offender recidivism was intended to summarize what is

known regarding the efficacy of sex offender specific

treatment, particularly as it compares to no treatment.

The authors' conclusions regarding the current status

of outcome research in the sex offender field were

generally quite negative. They found that methodology

was very poor in general, and they specifically noted

that treatment programs being studied were rarely

described clearly enough to determine why a particular

treatment worked or did not work. Furthermore, the

authors concluded that there was as yet no evidence

that treatment reduces sex offenses in general. Among

their conclusions was the need for developing short--

term data from which long-term data can be predicted

(Furby et al., 1989).

In the absence of empirically based amenability

criteria, treatment programs are nonetheless forced to

institute limits regarding eligibility and amenability

in hopes of using the treatment resource most

effectively and efficiently. In Faye Honey Knopp's

seminal work, Retraining Adult Sex Offenders: Methods

and Models, (1984) the problem of treatment candidate

selection is examined along with other key issues. The

majority of treatment programs for sex offenders use

some set of admission criteria to screen potential

candidates (Knopp, 1984). Residential treatment

programs continue to be relatively costly and limited

in availability; community based treatment programs

must consider their potential legal liability when

developing admission criteria and routinely exclude

violent offenders. Consequently, each setting has

evolved a set of eligibility criteria. These criteria

may be determined by law as in California and Florida,

or they may be derived from a consensus of program

officials or expert clinicians (e.g., Task Force of the

National Adolescent Perpetrator Network), tradition,

program design or multidisciplinary committees as in

Dr. Groth's now defunct program in Connecticut. The

members of the Governor's Task Force on Sex Offenders

and their Victims in their 1984 Report on Treatment

Programs for Sex Offenders regarding the state operated

sex offender treatment programs stated:

At present we are unable to pick out in advance
those individuals who are most likely to reoffend
unless treated and who will also respond well to
treatment. A high priority has to be given to
research to find methods of picking out those
offenders whose treatment will best serve the
needs of the community. (1984, p. 20)

Basic Assumptions, Techniques, Settings,
Populations Treated (Focus), Research
findingss and Related Literature

The various debates concerning the use of

outpatient vs. inpatient strategies, chemical vs.

behavioral vs. psychodynamic models, specialized vs.

generic treatments, incarceration vs. community

placement, with or without treatment, and determining

who is amenable for treatment and who should be

excluded remain unresolved in the sex offender field.

To date, researchers have offered very little in terms

of resolving any of these questions; they remain to be

answered by current (Marques, Day, Nelson, & Miner,

1990) and future researchers. However, in spite of

clinical imprecision, these questions are being

resolved routinely by bureaucrats and politicians

around the country (Knopp, 1984).

In a 1988 publication, Nationwide Survey of

Juvenile and Adult Sex Offender Treatment Programs and

Models (Knopp & Stevenson), the authors chronicle the

enormous growth in the number of sex offender treatment

programs in America as well as the variety of treatment

approaches offered. The report, based on a survey of

1002 respondents, elicited updated information

regarding sex offender treatment programs from a prior

(1986) survey. Additionally, a second part, including

responses from 574 adolescent and adult sex offender

treatment providers, elicited information regarding

"perceived program models, 43 treatment modalities, and

3 modes of treatment delivery." Responses from the

survey were structured so as to provide information

regarding the setting in which the treatment was

provided as well as the treatment models and modalities

employed. In the study a total of 429 adult

specialized sex offender treatment programs were

identified. Residential programs accounted for 28% of

all services (119); of these, 89% (106) were public

(eg. state funded) and 11% (13) were private (eg.

requiring payment from the client). It was found that

73% of residential programs were housed in prisons

(87), 16% were housed in mental health facilities (eg.


NFETC) (19), and 11% were located in private facilities


Almost three quarters of the specialized treatment

programs for adult offenders identified in the survey

were outpatient, community based (310 or 72% of all

adult services). Of these, 62% (191) were private

services, 31% (97) were located in mental health

centers, 4% (13) were court related services, and 3%

(9) were community based prison related services.

Residential Treatment Model

The 1988 Nationwide Survey conducted by Knopp and

Stevenson reported on "selected treatment methods" used

by respondents. The seven methods reported were family

therapy, peer-group treatment, "thinking errors,"

behavioral methods, aversive conditioning, Depo-Provera

(synthetic female hormones), as well as one evaluation

tool, the penile transducer or plethysmograph. Peer-

group treatment was the most widely used, being the

preferred method in 86% (371) of the 429 programs.

Family Therapy was included in 79% (341) of identified

adult sex offender programs, although only 54% (64) of

the 119 residential programs reported using it.

Behavioral methods were used in 65% (277) of the


identified programs. In descending order the frequency

of the remaining methods was Penile Transducer (26%),

Aversive Conditioning (24%) and Depo-Provera (18%).

Most often treatment programs offer one treatment

regimen which treatment candidates are required to

successfully complete. A diligent effort to define who

is amenable for a particular treatment model, however,

would be an important step in understanding individual

offender variables in terms of a given treatment design

and serve as a model for future research (Earls &

Quinsey, 1985).


Basic assumptions. Recently an integration of

cognitive-behavioral and psychodynamic models has been

suggested by several clinicians in the field as a means

of more effectively achieving positive long-term

results than cognitive-behavioral therapy programs have

thus far demonstrated by incorporating techniques

derived from psychodynamic therapies into a structured

treatment program. While as yet no outcome data exist

to specifically support this model, its supporters

believe it may be the answer to treating offenders with

more severe disorders who tend to be treated in

residential settings (Barnard et al., 1989; Shaw,

Hutchinson, & Longo, 1989).

Common techniques. Unique to the integrated

approach is the use of numerous theoretical models

within a coherent whole. Specifically cognitive-

behavioral modules are used in conjunction with

psychodynamic, experiential modalities such as gestalt

or psychosynthesis, as well as family systems to create

an integrated treatment program (Barnard et al., 1989).

Settings. Integrated treatment strategies are

found in both inpatient and outpatient programs for

both adult and juvenile sex offenders (Shaw et al.,


Populations treated. Integrated treatment

strategies have been used with all types of sex

offenders, both adult and juvenile.

Research evidence. Review of official arrest data

from Florida in 1988 and 1989 revealed that no

graduates of the integrated treatment program at NFETC

had been arrested for a sex offense up to the date of

the Annual Report to the Florida Legislature (Alcohol,

Drug Abuse, and Mental Health Program Office, Florida

Department of Health and Rehabilitative Services,



Basic assumptions. Therapies based on

psychodynamic models assume that sexual deviance is the

result of early perceived trauma and includes:

castration anxiety; reaction to seductive mother;

inadequate ego/superego; reenactment of sexual trauma;

confusion of aggressive and libidinal drives, and

narcissistic representation of self as child (Schwartz,


Common techniques. Psychodynamic treatment

techniques include insight oriented individual therapy,

including traditional analysis, and group counseling

including Gestalt therapy, Psychosynthesis, and

Psychodrama. These techniques are used to assist the

client/offender in becoming consciously aware of his

underlying motivations and subsequently developing new

ways of coping with his needs.

Settings. Psychodynamic treatment of sexual

offenders today is limited to individual therapy and

programs offered in several psychiatric hospitals. Due

to the length of time this treatment usually takes,

clinicians are frequently constrained by financial

considerations as well as time.


Populations treated. Psychodynamic treatment has

been used with a variety of sex offender populations

including rapists, pedophiles, exhibitionists, voyeurs

and other sexual deviants who may not be considered

offenders such as transvestites and various other


Research evidence. The majority of research

evidence for the efficacy of this approach has been in

the form of single case studies and has been less

evident in the last decade (Kilmann, Sabalis, Gearing,

Bukstel, & Scovern, 1982).


Basic assumptions. Cognitive-behavioral treatment

programs comprise the most widely accepted approach to

sex offender treatment today. This approach is based

on the assumption that behaviors are learned responses

and that individuals can learn new behavioral responses

and develop new beliefs to support those responses.

Cognitive therapies comprise a relatively new branch of

behavior therapy and focus on the thoughts, beliefs,

and fantasies which "drive" behavior. Deviant behavior

is often dependent upon distorted thinking (Marshall &

Barbaree, 1990).

Common techniques. Treatment programs variously

focus on preventing reoffense using some

combination/integration of Relapse Prevention,

cognitive restructuring, stress and anger management,

social and communication skills training, sex

education, and arousal reconditioning (Laws, 1990;

Pithers, 1990; Salter, 1988). Techniques employed by

these modalities include psychoeducational experiences

such as reading and viewing videotapes, modelling,

roleplaying and practice.

Settings. Cognitive-behavioral therapies are used

today in virtually all settings, whether the primary or

adjunctive therapy focus (Knopp et al., 1986; Knopp &

Stevenson, 1988).

Populations treated. Cognitive-behavioral

therapies have been used with virtually every sexually

deviant population treated; behavioral strategies such

as arousal reconditioning tend to be used for those

with evidence of deviant arousal or histories of

repetetive deviant behavior (Knopp et al., 1986; Knopp

& Stevenson, 1988).

Research evidence. Leaders in the outcome

research on cognitive--behavioral treatments, which

focus on arousal reconditioning (changing the focus of


arousal from a deviant to a non-deviant object or act),

include Abel and associates (1988), Earls and

Castonguay (1989), Laws and O'Neil (1981), and Marshall

and Barbaree (1988). William Pithers has designed and

implemented several studies using Relapse Prevention

based treatment programs in Vermont, including an

innovative approach which includes the training of

probation officers in Relapse Prevention (Pithers &

Cumming, 1989). He found that when both treatment and

follow-up probation supervision were based on Relapse

Prevention, recidivism in terms of probation violations

and rearrests for sex offenses was significantly

decreased. A proposal to compare a cognitive-

behavioral treatment program designed specifically for

sex offenders with a more generic treatment model based

on Sullivanian Interpersonal therapy was first approved

by the National Institute of Mental Health and then

rejected by the Human subjects review committee which

felt the control treatment was not likely to be

sufficiently effective (Laws, 1989, personal


In their review of outcome research in cognitive-

behavioral treatment therapies, Marshall and Barbaree

(1990) found promising results regarding use of these

modalities with exhibitionists and child molesters.

They concluded, however, that future outcome studies

must focus more on the specific changes induced by

treatment such as sexual preferences, social competence

and cognitive distortions to improve the ability of

clinicians to make predictions regarding likelihood of


Treatment Center Study/Setting


The sex offender program which is the setting for

this study was located at North Florida Evaluation and

Treatment Center (NFETC) in Gainesville, Florida.

Operating from 1976 until September, 1989, the program

was one of four treatment units within the treatment

center and one of two sex offender treatment programs

operated by the Florida Department of Health and

Rehabilitative Services (HRS).

The 63-bed MDSO unit was situated in three secure,

electronically controlled treatment buildings. Two

buildings housed 18 men in a two-pod floor plan and one

housed 27 men in a three-pod floor plan, with nine

individual rooms per pod. Each pod had a common area

for group or recreational activities which was

monitored visually and audibly by staff in a single

building control room. While the treatment program and

daily activities were facilitated by professional and

paraprofessional treatment staff, the secure perimeter

was maintained by uniformed security staff who were

available for backup in the event of a crisis or

emergency. Staff maintained 24 hour supervision.

Treatment Program Rationale

In 1985 the treatment program was redesigned to

reflect a multi-modal or integrated approach to sex

offender treatment, based on the belief that sex

offenders develop their deviant urges and behaviors for

a variety of reasons (Finkelhor, 1988).

The basic philosophy underlying the comprehensive
treatment program offered at NFETC is that sex
offenders develop their deviant behavior through
multiple and diverse ways and consequently require
a variety of treatment approaches to alter this
aberrant behavior. (Barnard et al., 1989, p. 126)

This model, consistent with the multi-factor

theory of offender etiology espoused by David Finkelhor

(1988) and heralded as a model for comprehensive

residential treatment, combined a cognitive-behavioral

treatment program, milieu therapy, and experiential

therapies designed to interface with each other

(Barnard et al., 1989). For example, an offender may

have been enrolled in Role Play (offense reenactment)


module where he and eight other offenders would reenact

their offenses in group in videotaped sessions. Later,

the offender would review the videotape of the role

play with his "Pod group" and primary therapist (i.e.,

case manager).

Treatment Program Format

Sex offenders participated in an initial screening

while in prison and were placed on a waiting list for

transfer to one of the two treatment facilities if

found eligible. Once transferred to NFETC offenders

underwent a 4-day period of evaluations while being

maintained in relative isolation to prevent treatment

from beginning until after the pretreatment evaluation

process was complete. This evaluation included a

Psychiatric interview, a computerized psychosocial

assessment including a plethysmograph assessment of

arousal patterns (Barnard, Robbins, Tingle, Shaw, &

Newman, 1987), and a WRAT-R. Following this battery

the offender participated in an 8-week "Evaluation"

phase which included several structured modules

(Criminal Thinking Errors, Psychosexual Process), other

structured activities including production of a

"Lifeline" and an "Offense Description," "Role Play" of

Offense, and unstructured group experiences at least

several times per week. At the end of this 8-week

Evaluation phase the offender was interviewed by the

multidisciplinary treatment team (MDT) which consisted

of the Unit Director, the 7 unit professional

therapists, the Unit Psychaitrist, the Unit Health

Coordinator (nurse), and available paraprofessional

staff. During this "staffing" the offender's response

to treatment was reviewed in terms of the following:

(a) his willingness and ability to actively participate

in the treatment components of the evaluation phase

such as lifeline creation, role play of offense and

thinking errors module (response to treatment); (b) his

willingness and ability to abide by program rules; (c)

his presentation in the staffing in terms of his

willingness and ability to describe and "own" his

offense, and his understanding of the precursors and

subsequent impact of his offense; and (d) a

psychological evaluation was also reviewed in this

staffing. Two decisions were possible at the

conclusion of the staffing. Based on consensus, the

team could decide to (a) accept the offender into the

"Treatment" phase (approximately one year) or (b)

reject the offender and return him to the custody of

the Department of Corrections (DC) as "not amenable for

treatment." A third decision, very rarely used, was

that in special cases (such as offenders with perceived

high motivation but marginal abilities) the evaluation

phase could be extended for another month.

During the treatment phase, offenders' progress in

treatment was reviewed at least monthly; offenders were

staffed formally at the end of one year, by which time

they were expected to have completed the modules, when

they were presented for completion of treatment, or

when either the offender or his primary case manager

requested a "termination" staffing, which could occur

at any time during the treatment process.

The treatment program was divided into three

stages, each with a set of specific activities and

expectations for the resident. In Stage I, the

evaluation phase, the offender was sequestered in

relative isolation until he completed the entire

psychosocial/psychosexual pre-test evaluation which

generally took from two to four days. Thereafter, he

was introduced to the other residents of the

intake/evaluation/advanced resident building and

expected to learn the rules of the program, attend pod

group two mornings per week, complete several standard

assignments including Lifeline, a structured

autobiography, offense descriptions for each separate

victim and type of offense he committed, a role play or

reenactment of either his instant offense or an

alternate agreed upon in advance by him and his

therapist, and psychosexual disorder process diagrams;

these had to be approved by both his therapist and his

pod group. In addition, the offender attended several

modules including the Criminal Thinking Errors module

facilitated by advanced residents, the Psychosexual

Disorder Module, Clinically Standardized Meditation

Module, resident government meetings which included

pod, building and unit functions, and special groups

called by staff or residents. The resident was

expected to meet at least once per week with his

primary therapist, his resident activity monitor (RAM),

a paraprofessional assigned to him to assist in the

completion and review of his "homework" assignments,

and his Triad members. Triads consisted of one

advanced resident and two residents in the evaluation

phase; triad members were required to be available to

each other at all times for support as well as for


At the end of the 8-week evaluation phase the

offender was interviewed by the MDT. During this


staffing the resident's response to treatment was first

reviewed for the group, usually by the primary

therapist, as well as results of a psychological

evaluation. After this review, the resident was then

interviewed, and his ability to describe his offense

without minimization, blaming or denial, his

understanding of his offense cycle including criminal

thinking errors he employed and any special issues or

questions pertaining to his amenabiilty were assessed

by the team. After the resident was dismissed from the

staffing, the team would complete a discussion of the

resident and vote on whether to accept or reject him

and return him to the department of corrections.

An offender who was accepted into Stage II of the

treatment program was expected to be enrolled in at

least three modules at any given time, each meeting

once per week, his pod group twice per week, structured

recreational activities including art, music, sports,

computer or video training, education classes for those

without a high school diploma, work program and

resident government activities. Stage II generally

lasted for eight to ten months, and most residents

completed this phase around one year from their date of

admission. At the end of phase II or one year from

date of admission, whichever came first, the resident

was staffed again. The staffing was basically the same

as the admission staffing except that during the

staffing the resident's progress through the modular

program was reviewed both in terms of modules completed

(grades) and his ability to describe how he could use

what he was learning to prevent future reoffenses and

other dysfunctional thinking and behavior. A report of

his progress was forwarded to the department of

corrections resultant from this staffing.

Stage III of the treatment program involved

retaking modules in areas where the offender appeared

weak (e.g., social skills), participating in pod

groups, resident government, Triad groups, facilitating

certain modules and developing release plans and

strategies. A final staffing was held when the MDT

determined that the resident had either maximized

benefits, made significant progress or had successfully

completed the program. This staffing was generally

structured around a review of progress including

available post-tests, recommendations from his primary

therapist and RAM, and an interview focused on release

planning, relapse prevention, and victim empathy.

Treatment Program Components

The following is a brief description of the

structured module components of the treatment program.

Residents were expected to rotate through these

modules, taking approximately three at any given time,

until all were successfully completed. Each module

lasted for approximately twelve sessions and was co-

facilitated; each session lasted from 1 1/2 to 2 1/2


The core treatment philosophy was presented

through the Empathy Development/Trauma Work component.

This set of modules was designed to increase the

offender's awareness of the impact of his offense on

both perpetrator and victim, assist the offender in

defining the purposes served by commission of the

offense, and promote the development of empathy for

victims. In the Role Play, a group experience, the

offender had the opportunity to act as both the

perpetrator of his own crime and as the victim in the

offense of a peer during the course of the module.

Each session of the module could be videotaped for

subsequent review by the offender.

A traumatic events component provided the offender

with specific skills to resolve traumatic events in

personal history in order to decrease resultant

cognitive distortions, emotional distress and

maladaptive behavior patterns. This was accomplished

through a series of therapist-facilitated structured

learning experiences and practical exercises

experienced in the "safety" of the therapeutic

environment using any number of modalities including

Gestalt Therapy and Neuro-Linguistic Programming.

A sensitivity training component assisted the

offender in developing an awareness of the connection

between body and mind, to facilitate their integration,

and to sensitize the offender to his own affective

experience and the experience of others, through the

use of therapist facilitated Bio-energetic, Gestalt and

structured training exercises.

A number of modules were designed to train the

offender in more adaptive life skills. These included

the following:

(a) The Relapse Prevention module was designed to

prepare the offender to take responsibility for his

behavior, particularly as it relates to his sex

offending, by identifying specific high risk situations

where offending is likely to occur, predicting their

occurrence, and acquiring specific coping skills and

developing action plans based on those skills to

prevent reoffense.

(b) The Arousal Reconditioning module included one

or both of the following components: Covert

Sensitization pairs aversive imagery and associated

negative affect with the antecedents to offending

behavior through directed, taped sessions; Self-

Administered Satiation focuses on decreasing deviant

arousal through the use of extended verbal repetition

of deviant fantasies in taped sessions with or without

concurrent masturbation in a controlled setting.

(c) The Cognitive Restructuring module, based on

Rational Emotive Therapy, was designed to directly

challenge the irrational beliefs which support the

offender's deviant behavior and to replace them with

beliefs grounded firmly in reality. This module helps

the offender develop responsibility for his own

experience and become more effective in relationships

with peers.

(d) The Stress Innoculation/Anger Management

module was designed to teach offenders how to better

cope with their anger and stress and to develop more

socially acceptable and personally effective

expressions of anger and stress coping responses.

(e) The Criminal Thinking Errors module, adapted

from the work of Yochelson and Samenow (1977), assists

the offender in identifying and replacing distorted

cognitions which support ongoing patterns of offending


(f) The Social Skills Training module was designed

to help the offender develop clearer, more effective

verbal and non-verbal communication skills, become more

assertive and develop socially appropriate and

effective hetero-social skills including dating,

courting and sexual interaction.

(g) The Healthy Sexuality/Sex Education module was

designed to improve the offender's knowledge of human

sexuality and to decrease distorted perceptions or

beliefs about sexuality.

(h) The Life Skills Training module could be

adapted on an individual basis depending on the

specific needs of the offender and the setting in which

the treatment is occurring.

(i) The Substance Abuse module was designed to

help the offender recognize substance abuse problems as

well as provide general information on the causes and

effects of substance abuse.

(j) The Psychosexual Disorders module presents an

introduction to the concept of the core process of the

psychosexual disorder which serves to provide a

foundation for integration of later treatment

experience into a comprehensive and cohesive model for

ongoing recovery.

Target Population

The sex offender unit at NFETC is a treatment
program for men who have been convicted of a
sexual offense and sentenced to a prison term.
The MDSO population is comprised of child
molesters, incest offenders, exhibitionists and
rapists who volunteer for treatment and pass
screening by a prison psychologist and a
professional staff member of the Sex Offender
Treatment Unit. Final selection into the
treatment program is determined at the end of an
eight week clinical evaluation period.
The following criteria are used in the screening
process for accepting sex offenders into the
program: (1) there must be evidence of a
psychosexual disorder; (2) the inmate must
volunteer for treatment; and (3) the inmate must
accept responsibility for the crime. Inmates are
automatically excluded from MDSO treatment if they
present one of the following characteristics: (1)
have a sentence in excess of 15 years or less than
18 months; (2) show significant evidence of a
major mental illness (e.g., psychosis); (3) have
murdered their victimss; (4) are actively
appealing their conviction or sentence; or (5)
have significant medical problems which may be
exacerbated by the stress of the treatment
program. (Barnard et al., 1989, pp. 122-123)

Treatment Amenability Factors

There is little question that the theoretical

approach employed by the treatment program is likely to

have an effect on who is considered amenable for that

treatment. Characteristics of offenders likely are

important indicators of amenability for treatment.

These can be divided into four categories: (a)

intellectual factors, (b) personality factors, (c)

offense factors, and (d) demographic factors.

Intellectual Factors

Several authors have noted that a particular model

requires some minimum reading skills or the ability to

integrate material into daily life which is frequently

referred to in the literature (Barnard et al., 1989).

Some form of intelligence and/or performance evaluation

is routinely considered in amenability decisions

although there is little consensus in the literature as

to what the ideal should be. A residential sex

offender program in California requires a minimum of 80

IQ for consideration in treatment (Marques, et al.,

1990). Summarizing his views on the ideal treatment

candidate, Smith (1988) states that offenders should

have normal intelligence.

Personality Factors

Marcus (1971), in listing criteria which would

negatively affect amenability, included delinquent acts

between the ages of 8 and 13, interrelated criminality

with sexual offenses, lack of concern for victim.

Numerous other clinicians and researchers have offered

sets of eligibility criteria. For example, co-

directors of a treatment program for sex offenders in

New York report that assessment is particularly

important to assess personality characteristics. They

rule out offenders who are "actively psychotic,

predominantly antisocial, or heavy substance abusers"

(Travin, Bluestone, Coleman, Cullen, and Melella,


Offense Factors

Although few researchers have set out to predict

amenability based on treatment outcome measures, the

most notable data derive from measures within a long-

term study. Recently, Abel et al. (1988) published a

study predicting response to outpatient treatment of a

group of child molesters. The two outcome variables

for this study were dropout (initiated by either the

offender or clinician), a short-term outcome variable,

and recidivism (self-report of reoffense as determined

through structured interviews at six months and one

year post-treatment), a long-term treatment outcome

variable. There were 192 subjects, all pedophiles, who

were divided into 19 groups. Groups received the

treatment modules in varying orders in order to test

the efficacy of each module. The treatment program

consisted of thirty 90-minute weekly group sessions.

Each module or treatment component consisted of ten

sessions. The three treatment components were (1)

decreasing deviant arousal, (2) sex education/sex

dysfunction and cognitive restructuring, and (3) social

and assertiveness skills training.

The first dependent variable, dropping out of

treatment, can be viewed as a short-term outcome and a

measure of amenability. Approximately one third of the

subjects either dropped out (88%) or were expelled

(12%) prior to completing the thirty sessions. A

series of t-tests were conducted to determine if

significant differences on a number of demographic and

offense variables could be found between the group

which completed treatment and the group which did not.

Interestingly, most of the demographic variables (age,

race, social class, marital status, education,

employment status, and religious preference), referral

source as well as degree of motivation for seeking

treatment, lifetime reported number of pedophilic acts

prior to entering treatment, and self-reported current

ability to control urges all failed to discriminate

between the two groups. However there were three

characteristics which significantly differentiated

those who dropped out from those who completed

treatment: "(1) the amount of pressure the subject was

under to participate in treatment (p <.05), (2) the

diagnosis of an antisocial personality disorder (p

<.01), and (3) the lack of discrimination in the choice

of sexual victim or paraphilic act (p <.0001)" (Abel

et al., 1988). The dropout group reported greater

pressure to participate in treatment, was more likely

to have a diagnosis of antisocial personality disorder

and was less discriminating in the choice of sexual

victim or paraphilic act. A discriminant function

analysis was performed to estimate the degree to which

dropping out of treatment could be predicted using

variables that significantly differentiated the two

outcome groups. Using the variables stated above, the

discriminant function analysis could correctly classify

72.4% of subjects entering treatment as to whether they

would drop out of treatment.

Demographic Factors

Researchers have begun to test assumptions of

clinicians regarding demographic variables likely to

affect treatment amenability. One of the most

promising is marital status. Others which have not

received empirical support are age, race, socio-

economic status, education, employment status,

religious preference and lifetime number of reported

offenses (Abel et al., 1988).

Independent Variables

Reading Ability

As has already been suggested, predictors of

amenability are likely to be program specific. Perhaps

the most common feature of contemporary sex offender

treatment programs is the requirement that offenders

learn information in treatment modules, understand this

information, and integrate it in a way such that they

can effectively use what they have learned to prevent

engaging in a reoffense at some later date. While IQ,

as measured by the WAIS-R, is the most common measure

of ability, reading achievement, as measured by the

WRAT-R, is a more pragmatic measure, easier to

administer and score than the WAIS-R, yet positively

correlated with the WAIS-R (Margolis, Greenlief, &

Taylor, 1985; Spruill & Beck, 1986; Cooper & Fraboni,


As has been noted, numerous references are to be

found in the literature suggesting that intelligence

and reading achievement are predictors of successful

completion of treatment. An exception to this is

Marcus (1971) who wrote that high IQ offenders should

be screened out of treatment programs because of the

likelihood that they would manipulate the staff and

thereby avoid the full impact of the treatment process.

This aside, the predominant notion in the field is that

it takes some finite and measurable degree of

intelligence to complete a treatment program

adequately, and this has been suggested by numerous

authors (e.g., Marshall & Barbaree, 1988). But it is

clinical experience which overwhelmingly supports the

contention that the lower functioning offenders will

take longer to complete treatment and may not, in fact,

be amenable to cognitive-behavioral or insight oriented

therapies. No doubt for this reason virtually every

treatment program for sex offenders has an IQ and/or

reading achievement cut off as part of the admission


In a chapter titled "Clinical Assessment of Sex

Offenders" the author comments, "A lack of intellectual

ability might preclude certain verbal therapies or

imagery-based behavior therapy" (Dougher, 1988, p. 78).

Smith, of the National Academy of Corrections, in

discussing treatment of mentally ill and mentally

retarded sex offenders, further stated:

The experience of programs that integrate low
functioning or mentally ill men into confrontative
or cognitively oriented programs has not been
promising. (1988, p. 35)

Interestingly, as mentioned above, Marcus (1971) took a

minority position that high IQ is an undesireable trait

because it could be used by the offender to manipulate

the staff or therapist and consequently avoid the full

impact of the treatment process. Nonetheless, his

notion further underscores the need to study the

corellation of this variable with treatment outcome.

Antisocial Personality Disorder

Although only appearing briefly in the research

literature as a factor predicting dropout (Abel et al.,

1988), a diagnosis of Antisocial Personality Disorder

(APD), determined primarily by history, is considered

by many clinicians to suggest a poor prognosis for

lasting participation in treatment programs or

resultant meaningful change.

Among the criteria of APD which clearly would

predict problems in treatment are the following:

Onset before age 15 as indicated by a history of three

or more of the following before that age: truancy;

expulsion or suspension from school; persistent lying;

thefts; vandalism; school grades markedly below

expectations in relation to estimated or known IQ;

chronic violations of rules at home and/or at school;

initiation of fights. At least four of the following

manifestations of the disorder since age 18: inability

to sustain consistent work behavior; failure to accept

social norms with respect to lawful behavior;

irritability and aggressiveness as indicated by

repeated physical fights or assault; failure to plan

ahead, or impulsivity; disregard for the truth as

indicated by repeated lying; recklessness. (DSM III,


Any one of the above mentioned criteria, if it

were detected repeatedly or, in cases of extreme

examples, even once, in a treatment program would

likely result in the offender being terminated from

treatment (North Florida Evaluation and Treatment

Center, Sex Offender Unit, 1988b); if not detected,

these behaviors would be likely to disrupt the

treatment milieu and cause the offender to benefit less

from treatment. Consequently, many programs routinely

exclude these individuals, particularly those in

community, outpatient settings. In programs designed

to treat incarcerated offenders, excluding those

offenders with APD is more difficult because they

comprise a large percentage of treatment referrals and


Schwartz describes a treatment program at the

Chittenden Correctional Facility in South Burlington,

Vermont that selects offenders

with demonstrated histories of prosocial
behaviors, and those who don't have a number of
circumstances which threaten the individual's
sense of self control, e.g., alcoholism.
Candidates must accept responsibility and
acknowledge the harm done. Sadistic offenders are
excluded. (1988b, p. 47)

In Abel's landmark study the authors commented:

It should also not be surprising that individuals
with antisocial personality are more likely to
drop out of treatment, since a characteristic
frequently seen in such personalities is
manipulativeness of others to get what they want.
In this situation, admission to a treatment
program was probably helpful to them at the
moment; and as soon as they had satisfied someone
by their entrance into treatment, they quickly
terminated the treatment program. (Abel et al.,
1988, p. 229)

Even if the antisocial personality disordered

offender did not terminate himself from treatment it

appears likely that he would be frustrated by the

control the residential program exerted over him;

instead of working diligently to make meaningful


changes in himself, he would focus on manipulating the

staff and his peers, on avoiding the true thrust of

treatment in ways he had developed in his past such as

missing classes, failing to complete homework

assignments, or faking illnesses. Consequently, it

could be predicted that he would not do well in

progress staffing, either from poor treatment

participation or due to program rule infractions, the

other serious behavioral manifestation of APD. The

individual with APD does not generally respect rules

and follows them begrudgingly; he focuses on not

getting caught and in a residential program might be

likely to cheat on tests, copy homework assignments

from others, steal, threaten, fight, use drugs,

manufacture alcoholic beverages and other behaviors

generally against program rules and often requiring

termination from the treatment program. If any of

these behaviors were discovered, they would negatively

impact on perceived treatment outcome by the MDT.

"Career" criminals and others with entrenched

antisocial personalities rarely respond positively to

treatment (Dougher, 1988). Therefore, it is reasonable

to assume that a diagnosis of Antisocial Personality

Disorder would be positively correlated with either

dropout or unsuccessful completion of treatment.

Degree of Discrimination in the Choice of Offense and
Victim Type

This variable has been included for testing for

several reasons. Abel and associates (1988) found, in

a sample of outpatient pedophiles, that this variable

predicted both poor response to treatment in the form

of dropout as well as recidivism using self-report of

reoffense as the outcome measure. In fact, Abel found

that, "All 30 offenders who dropped out during the

first 10 weeks of treatment had committed acts against

both males and females and against both children and

adolescent victims. More than half of the subjects who

dropped out before the end of the 30 week treatment

showed this same multiplicity of diagnoses, which was

significant at the p <.0001 level" (p. 227).

It was also found that offenders who had committed

both "hands-on" assaults of children as well as "hands-

off" assaults (such as exhibitionism or voyeurism) had

higher rates of dropping out than those who had

committed only one type or the other (p <.0001).

When the authors combined these characteristics they

found that 89.9% of the offenders who dropped out of

treatment had committed acts against males and females,

children and adolescents, both "hands-on" and "hands-

off". These findings were significant at the p <.0001

level. When relationship to the victim, described as

incest and nonincest, was considered with the above

factors the percentage of those dropping out remained

high (88.1%, significant at the p <.0001 level).

Concluding his review of the findings related to

dropping out, Abel et al. stated:

Combining all the mutiplicities of target
characteristics, it was found that 35 subjects had
committed pedophilic acts against males and
females, children and adolescents, and incest and
nonincest victims using hands-on and hands-off
molestation. Almost all of them (32 subjects, or
91.4%) dropped out of the treatment program. Of
the 157 subjects who did not report such
multiplicity of behaviors before treatment, only
22.3% dropped out of the treatment program. (Abel
et al., 1988, p. 228)

In interpreting their data the authors suggest

that a multiplicity of offense and victim interests and

behaviors may make behavioral techniques aimed at

decreasing deviant interests ineffective. They

further suggest that the individuals with the greater

diversity of interests and behaviors may be

qualitatively different from other offenders, more

deviant or suffering from a more serious "dose" of the

disorder which is more resistant to treatment induced

change and therefore more likely to resist treatment in

the form of dropout or unsuccessful completion and more

likely to recidivate.


How age affects amenability continues to be a

clinical issue in sex offender treatment; there are

several components to this issue. Numerous studies of

incarcerated offenders have shown that the majority of

individuals entering prison for violent crimes are

between the ages of eighteen and thirty-five; a sub-

group of this population is comprised of rapists

(Scully, 1990). The treatment of adolescent sex

offenders is supported in part by the assumption that

early intervention is likely to stop a pattern of

behavior before it becomes deeply entrenched. But

recent studies have shown that most sex offenders, by

their mid-twenties, are already deeply entrenched in

their deviant patterns. Many of these offenders remain

relatively immature, defiant and may be resistant to

change. The maturing process has long been assumed by

clinicians to be an ally of treatment. Thus, once

offenders have reached adulthood, their amenability for

treatment is likely to improve as they get older; as

they mature they may be more likely to consider the

consequences of their behavior and be more willing to

seek alternatives to their deviant interests.

Marital Status

Several researchers have found that a stable

marital relationship is a predictor of successful

completion of treatment as well as nonrecidivism.

Schwartz (1977) found that intact marriage was a

predictor of nonrecidivism in a sample of sex

offenders, including rapists, being treated in the

community; Abel and his associates (1988) found that

being married was a predictor of successful completion

of treatment as well as nonrecidivism for the

pedophiles in their outpatient treatment program.

Instant Offense

The instant offense, or the one for which the

offender is currently incarcerated, is often used as

identifying data; an offense for which the offender was

convicted is not dependant upon self-report for

verification. Perhaps one of the most commonly held

clinical assumptions in the sex offender field is that

among rapists, child molesters (extrafamilial

pedophiles) and investors (intrafamilial), rapists will

have the poorest prognosis for successful treatment,

incest offenders will have the best prognosis, and

child molesters will fall somewhere in between.

Outcome data from incest offender treatment programs

are frequently disregarded by clinicians working in

secure, inpatient facilities where, at best, incest

offenders comprise perhaps one-third of the treatment

populations and where they are considered the easiest

clients. Clinicians in these programs, recognizing the

need for programs capable of treating rapists and child

molesters, grudgingly accept these clients inspite of

their poor prognosis, recognizing the need to decrease

their dangerousness, while, at the same time, fearing

the negative publicity should one of these offenders


Treatment Outcome Categories

Very few sex offenders volunteer for treatment

without having been coerced or ordered into it by some

form of the judicial or corrections systems.

Frequently, reports on progress and ultimately

prognosis in terms of prediction of reoffense are

required by these agencies and information and opinions

provided often have a significant impact on judicial

and corrections decisions concerning the offender. As

a result of this and the fact that a significant number

of sex offender treatment programs are operated by

agencies with public funds, formal, documented

decisions concerning amenability, prognosis, progress

and dangerousness are frequently required of clinicians

in this field.

Termination at the End of the 8 Week Evaluation Phase-
-Group (a)

The criteria for membership in group (a) are as

follows: discharged during or at the conclusion of the

evaluation phase of treatment either at the resident's

request or as a result of a decision by the MDT and

officially designated "not amenable for treatment"

(those discharged due to severe medical problems would

be excluded from the study, since they should have been

excluded from the waiting list and were inappropriately

referred for treatment); not accepted into the

treatment phase at the acceptance staffing attended by

the mdtt after the 8-week evaluation phase.

Termination During the Treatment Phase with a Poor
Prognosis--Group (b)

The criteria for membership in group (b) are as

follows: accepted into the treatment phase at the

acceptance staffing attended by the MDT after

successfully completing the 8-week evaluation phase;

subsequently determined to be no longer amenable for

treatment at an official termination staffing for one

or more of the following reasons: inappropriate

behavior such as repeated threats of violence, an act

of violence, sexual misconduct or other serious rule

violation; failure of treatment modules; failure to

follow treatment plan; resident request for termination

from the treatment program; a combination of the above,

or determined by the MDT to have made "no progress" or

"limited progress."

Termination at the End of the Treatment Phase with Good
Prognosis--Group (c)

The criteria for membership in group (c) are as

follows: accepted into the treatment phase after

successfully completing the 8-week evaluation phase;

subsequently determined to have made "significant

progress" or to have "successfully completed" the

treatment program at the time of discharge.

Criteria for a rating of "significant progress"

included successful completion of most or all treatment

modules; demonstration by observable behavior of

significant integration of treatment material into

daily behavior; ability to describe the causes and

effects, long- and short-term, and possible

interventions of the sex offending behavior. Criteria

for a rating of "successful completion" included being

accepted into the treatment phase after successfully

completing the 8-week evaluation phase; successful

completion of all treatment modules; demonstration by

observable behavior of consistent integration of

treatment material into daily behavior; ability to

describe the causes and effects, long- and short-term,

and possible interventions of the sex offending

behavior; having an approved release plan.


Although much work has been done in recent years

developing strategies to treat sex offenders, there

remains little empirical evidence in the research and

practice literature regarding the effectiveness of

these strategies and even less regarding the more

specific question of which types of treatment are

effective with which sex offenders. In this chapter

the research and practice literature has been reviewed,

and common assumptions held by clinicians have been

discussed. A set of offender variables have been

described which are examined in this study.


This study was designed to determine if six

offender variables, assessed during the first 8 weeks

of a residential sex offender treatment program, were

correlated with membership in one of three treatment

outcome groups. The offender variables consisted of

(a) reading level, (b) presence of antisocial

personality disorder, (c) degree of discrimination in

the selection of offense and victim type (operationally

defined as total deviance score), (d) age, (e) marital

status, and (f) instant offense. The treatment outcome

groups consisted of (a) offenders who were discharged

at the end of eight weeks as nonamenable, (b) offenders

who were judged amenable for treatment at the end eight

weeks but were subsequently discharged prior to

completing treatment with a poor prognosis, and (c)

offenders who were judged amenable for treatment at the

end of eight weeks and were subsequently discharged

after completing the program with a good prognosis.

This chapter consists of sections describing the

research design, population and sample, procedures for

treatment and data collection, instrumentation,

hypotheses, data analyses, and limitations of the


Research Design

Criterion Variable

In this study an ex post facto design was used to

evaluate the interrelationships among six offender

variables and three treatment outcome groups: (a)

offenders judged nonamenable for treatment during an

evaluation phase of a residential treatment program,

(b) offenders judged amenable for treatment during an

evaluation phase, but later discharged with a poor

prognosis, and (c) offenders judged amenable for

treatment during an evaluation phase and subsequently

discharged with a good prognosis.

The criteria for membership in group (a) were

discharged during or at the conclusion of the

evaluation phase of treatment either at the resident's

request or as a result of a decision by the MDT, and

not accepted into the treatment phase at the acceptance

staffing attended by the treatment team after the 8

week evaluation phase and officially designated "not

amenable for treatment." The criteria for membership


in group (b) were accepted into the treatment phase at

the acceptance staffing attended by the MDT after

successfully completing the 8 week evaluation phase and

subsequently determined to be no longer amenable for

treatment at an official termination staffing for one

or more of the following reasons: inappropriate

behavior such as repeated threats of violence, an act

of violence, sexual misconduct or other serious rule

violation; failure during treatment modules; failure to

follow treatment plan; resident request for termination

from the treatment program; a combination of the above;

or determined by the MDT to have made "no progress" or

"limited progress." The criteria for membership in

group (c) were being accepted into the treatment phase

after successfully completing the 8 week evaluation

phase and subsequently being determined to have made

"significant progress" or to have "successfully

completed" the treatment program at the time of


Criteria for a rating of "significant progress"

included (a) successful completion of most or all

treatment modules, (b) demonstration by observable

behavior of significant integration of treatment

material into daily behavior, and (c) ability to


describe the immediate and long-term causes and impact

of the sexual offending and to demonstrate knowledge of

approved methods for preventing future sex offending

behavior. Criteria for a rating of "successful

completion" included (a) successful completion of all

treatment modules, (b) demonstration by observable

behavior of consistent integration of treatment

material into daily behavior, (c) ability to describe

the immediate and long term causes and impact of the

sexual offending and to demonstrate knowledge of

approved methods for preventing future sex offending

behavior, and (d) an approved release plan.

Independent Variables

The six variables examined in relation to the

outcome variables were (a) reading level, reported as a

raw and converted score from the administration of the

Wide Range Achievement Test-Revised (WRAT-R); (b)

presence of an antisocial personality disorder,

documented by the primary case manager who was

responsible for integrating the consensus diagnoses of

the MDT; (c) degree of discrimination in the selection

of offense and victim type (total deviance score) based

on the documentation in the clinical summary of offense

history from official reports and self-report of the

offender during the evaluation phase; (d) age at time

of admission; (e) marital status, and (f) instant



The population from which the sample was drawn

consisted of 167 males who (a) were convicted of sex

offenses including rape, child molestation and incest,

(b) were incarcerated in the Florida prison system

between the years of 1985 and 1989, (c) were between

the ages of sixteen and seventy, (d) had a minimum of a

third grade reading level and a score of 70 or higher

on the WAIS-R, (e) had no organic, psychotic or other

major mental disorder, (f) volunteered for

participation in a Health and Rehabilitative Services

(HRS) residential sex offender treatment program, (g)

were placed on a waiting list for treatment, and (h)

were subsequently transferred to NFETC when they reached

the top of the waiting list.

Each male sex offender who was convicted and

sentenced to a period of incarceration with the Florida

Department of Corrections (DC) was required by statute

to be evaluated for possible participation in one of

the two HRS Mentally Disordered Sex Offender (MDSO)

treatment programs. Preliminary screening was

accomplished by DC psychological staff to identify

those offenders who were administratively qualified for

transfer to HRS for sex offender treatment and who

volunteered for that treatment. Such offenders were

subsequently interviewed by a joint screening team

composed of representatives of the DC psychological

staff and the staff of one of the HRS sex offender

treatment programs. The following criteria were used

in the joint screening process to determine potential

eligibility for treatment: (a) evidence of a

psychosexual disorder, (b) willingness of the offender

to volunteer for treatment, (c) demonstration by the

offender of sufficient motivation and capacity for

insight to indicate a possibility of successful

treatment completion, and (d) A score of 70 or above on

the WAIS-R or equivalent and WRAT-R of third grade or


Offenders who were determined to be potentially

eligible for consideration for treatment as a result of

this joint screening process were then placed on a

waiting list to await transfer to HRS custody for

treatment. This waiting list was prioritized according

to the length of time remaining on the offender's

sentence, with those having the least time being

afforded the earliest opportunities for treatment.

Those offenders with shorter sentences which precluded

a minimum of at least eight months treatment were

excluded from participation.

Resultant Sample

The resulting sample of 114 offenders consisted of

those individuals from the population described who had

not been accepted for treatment prior to the

establishment of the new treatment program in July,

1986 and had sufficiently progressed in treatment to

receive a completion rating at the time of the closing

of the program in June, 1989. Excluded from the sample

were offenders found not amenable for treatment, once

at the facility, due to criteria which would have made

them ineligible at the time of original screening such

as a chronic physical problem (e.g., heart disease) or

major thought disorder.

The sample of offenders whose records were studied

consisted of 114 male sex offenders who attended the

sex offender treatment program at North Florida

Evaluation and Treatment Center (NFETC) between March,

1986, and August, 1989. The mean age of these

offenders on admission to the treatment facility was

31.4 years of age with the youngest being 16 and the

oldest being 59. Of the 115 offenders, 47 (40.87%) had

been incarcerated for a most recent charge (instant

offense) of incest or intrafamilial child molestation.

Extrafamilial child molesters, those who victimized

children outside of the home, numbered 37 (32.17%).

There were 31 (26.96%) men incarcerated for rape.

Examining the marital status of the offenders revealed

that 40 (35.09%) were married, 39 (34.21%) had never

been married, 9 (7.89%) were separated, and 26 (22.81%)

were divorced. There were 27 (23.48%) black offenders,

87 (75.65%) white offenders, and one (0.88%) hispanic

offender in the sample.


Treatment Procedures

The residential treatment program from which data

for this study were collected was housed at North

Florida Evaluation and Treatment Center (NFETC), a

maximum security forensic mental health institution

operated by HRS in Gainesville, Florida since 1976. The

MDSO program, which comprised approximately one-third


of the institutional population, had a maximum capacity

of 63 beds dedicated to the treatment of convicted male

sexual offenders. The NFETC sex offender program

physical plant included three buildings, two of two

wings (called pods) housing offenders who had been

accepted for the treatment phase, and one of three pods

which included two pods of offenders in the 2-month

evaluation phase and one pod of advanced offenders

approaching successful completion of treatment. Each

pod was comprised of 9 offenders, each of whom was

assigned an individual room. Each of these offenders

had been screened and approved for transfer at one of

various DC institutions throughout Florida prior to

transfer to the MDSO unit.

The treatment program integrated three broad areas

of therapeutic focus: (a) assisting the offender in

replacing distorted or irrational thinking and

maladaptive behaviors commonly associated with the

psychosexual disorder (e.g., cognitive restructuring,

behavioral module); (b) training offenders in the

skills necessary to improve their interpersonal and

intrapsychic functioning (e.g., their communication

skills and stress/anger management); and (c) assisting

the offender in identifying and remediating impediments

to healthy functioning that were the result of early

trauma and/or lifelong acquired patterns of dysfunction

through corrective emotional experiences using

techniques derived from Gestalt Therapy,

Psychosynthesis and Neuro-Linguistic Programming as

well as modified aversive behavioral rehearsal therapy

(Knopp, 1984).

A basic assumption underlying this treatment

approach was that there is an endless variety of

combinations of behavioral, experiential and cognitive

antecedents that sex offenders describe in their

offense cycles. The idea that one treatment modality

would successfully ameliorate the problems of so

diverse a population seemed optimistic at best and

grossly inadequate at worst. Based on their treatment

experience, the staff concluded that the best strategy

for treating such a diverse population was to offer a

variety of treatment modalities, individualizing the

treatment by stressing some components over others

based on evaluations of the offender's behavior,

progress, and personality configuration.

A second concept underlying the treatment program

was that, whenever possible, treatment would be

delivered using a structured module format to


facilitate accurate replication for quality control and

future research. Topics addressed by structured

modules ranged from educational material such as sex

education to sensitivity training which included

experiential components. Although some modules were

appropriate for all offenders in the program,

individuals with extensive histories of abuse were

considered to require more emphasis on repairing the

effects of early trauma and more supportive therapy

than those with less abuse in their early history.

Preliminary assessment phase. Upon admission,

each offender was required to participate in a

preliminary treatment assessment phase. During this

phase each offender participated in an initial intake

interview with members of the MDT and then took a

battery of assessment instruments. Also during this

phase, each offender was sequestered in his assigned

room except when participating in an initial

comprehensive assessment battery.

Evaluation phase. The 8 week evaluation phase,

which began with the Computer Assisted Psychosocial

Assessment (CAPSA), was designed to begin the process

of treatment while testing the offender's ability and

motivation to participate successfully in the program


components. To this end this phase consisted of

learning the program rules, participating in structured

and unstructured groups on a daily basis, completing

chores, completing frequent structured assignments

designed to increase awareness of the offender's

offense cycle, and revealing to his therapist and group

members significant information from his past regarding

his development including his offenses as well as his

own traumas. Structured modules included criminal

thinking errors and offense cycle (central process).

During this phase each offender received a

psychological evaluation which was included in his

clinical record. Each offender's progress was reviewed

weekly by the MDT including the paraprofessional staff

person called a Resident Activity Monitor (RAM), who

tracked the offender's compliance with treatment plan

directives. At the conclusion of the evaluation phase

the offender was evaluated by the MDT in a formal

review known as the Acceptance Staffing where the

offender's progress, amenability for treatment as

predicted by the psychological evaluation, and the

offender's ability to respond to stressful questioning

were considered in determining his eligibility to

continue into the treatment phase of the program.


Treatment phase. The treatment program utilized a

therapeutic community to create an emotionally safe

environment within which to explore past traumatic

experiences, faulty learning, and inappropriate

behavioral reponses to stressful situations.

Responsible for this milieu were the 7 professional

therapists and 45 paraprofessional staff who managed

the three treatment buildings. This therapeutic

community included a resident government, peer-

facilitated groups and impromptu encounters providing

the background upon which this modularized program was

superimposed. Material learned in the modular program

was expected to be demonstrated in the milieu in new,

more appropriate behavior.

Additionally, the program included therapist-

facilitated pod groups which met twice weekly to focus

on here-and-now interpersonal conflicts and progress

using core program components such as "central process

of the psychosexual disorder," bodywork and meditation,

the offender work program, Alcoholics Anonymous, basic

adult education, evening study halls, structured

training groups, and art, music, and recreation


Treatment participants who had completed all the

treatment modules and had advanced to the integration

phase assisted in facilitating groups and also served

as primary facilitators for structured evening training

modules on a regular basis in the evaluation building.

Offenders provided support and facilitation to one

another throughout the treatment week in the form of

Triad Sessions (small group teams of three) and special

group attendance.

There was also a peer-facilitation/peer-evaluation

process interwoven into the structure of the overall

treatment program which was directed by the Offender

Government. Regularly scheduled, periodic reviews of

the offender's progress were conducted by offender

review boards comprised of elected peers, and the

results were presented to the offender and his case

manager as feedback.

Evaluation of each offender's progress in

treatment was conducted by staff in accordance with

standards mandated by HRS. Progress in treatment was

reviewed with each offender on a weekly basis during

the initial two month evaluation phase and biweekly

during the remainder of his treatment. Regular

progress staffing were scheduled and conducted

annually, and special staffing were scheduled as

therapeutically indicated on a case by case basis.

The treatment program at NFETC was structured to

be 18 months in duration. Although a few offenders

were able to complete all of the modules and

demonstrate integration of the skills taught into daily

behavior within as little as 15 months, most required

at least 18 months to successfully complete the

program. Regardless of the length of time an offender

spent in the treatment program, he received regular

feedback on his progress from the treatment team.

Data Collection Procedures

Data for the study were drawn from case files in

the Medical Records department of NFETC. Each

offender's record was reviewed to determine the WRAT-R

raw score and grade equivalent. These scores were

either documented in the offender's record from his DC

evaluation or in the psychological evaluation conducted


Also reviewed was the diagnosis. In order to

maximize reliability, the diagnosis in the initial

clinical summary was used because it documents a

consensus decision of the MDT. A zero (0) was scored

for no diagnosis or no features of antisocial

personality disorder. A one (1) was scored for a

diagnosis with features of antisocial personality

disorder, and a two (2) was scored for a diagnosis of

antisocial personality disorder.

The clinical summary(ies) were also reviewed to

determine the number of different victim types (age and

gender) as well as the number of different types of

paraphilic acts. If the information regarding victim

type and/or paraphilic acts was unclear in the summary,

a further review of the case file including

documentation of self-reports of previously unknown

offenses revealed by the offender during clinical

interviews or documented from other treatment

modalities was conducted. The clinical summary(ies)

was also reviewed to determine the resident's age upon

admission, marital status and instant offense.


This section contains descriptions of the

instruments used to assess the variables of interest in

this study. Included are the Wide Range Achievement

Test-Revised, the evaluation phase clinical summary,

and the NFETC clinical case record.

Wide Range Achievement Test-Revised (WRAT-R)

The Wide Range Achievement Test-Revised (WRAT-R)

will be used to measure reading level. The WRAT was


first developed and standardized in 1936 as an adjunct

to intelligence tests and measures of behavior

adjustment. Since its development it has been revised

five times and has been experimentally administered to

thousands of people from pre-school through adulthood.

Although the 1984 WRAT-R contains many changes

from the original, it remains basically the same in

form, content and commitment to remaining a simple

assessment instrument (Jastak & Jastak, 1979). The

thrust of recent revisions has been to make the WRAT-R

an up-to-date psychometric instrument using improved

methods for item analysis and scaling (i.e., Rasch


The WRAT-R provides data regarding the level of

reading skills in both raw score form and as equivalent

school grades. It provides a more direct measure of

ability than the WAIS-R with which it is, however,

highly correlated, particularly the verbal IQ and full-

scale IQ (Cooper & Fraboni, 1988; Margolis, Greenlief &

Taylor, 1985; Ryan & Rosenberg, 1983; Spruill & Beck,

1986). This high correlation holds for geriatric

populations (Margolis et al., 1985); mixed mental

health patients (Ryan & Rosenberg, 1983); and normals

(Cooper & Fraboni, 1988; Spruill & Beck, 1986). The

advantage of using the WRAT-R as a measure of ability

instead of the WAIS-R is that the WRAT-R is relatively

easy and quick to administer and is therefore more

likely to be available to most programs, regardless of

size and psychological resources.

Assessment of Antisocial Personality Disorder

The presence of Antisocial Personality Disorder or

other diagnoses were determined by an initial intake

interview, subsequent social history, information

revealed during the Evaluation Phase and a

psychological evaluation and were documented in the

Evaluation Phase clinical summary. The Evaluation

Phase clinical summary is a document developed by the

staff of NFETC for reporting of the results of the

eight week evaluation phase of the sex offender

treatment program and sent to the Department of

Corrections. It was the responsibility of the

offender's primary case manager to produce following

NFETC summary guidelines. It included information

derived from interviews, treatment procedures, and

previous documentation. Among its contents was a

consensus Axis II diagnosis opinion of the MDT

attending the 8-week evaluation staffing generally