A Structured Method of Assessing Dynamic Risk Factors Among Sexual Abusers With Intellectual Disabilities

Abstract

The nature and severity of dynamic risk factors among a group of 87 adult male sexual abusers with intellectual disabilities were examined as was the psychometric properties of a new scale designed to measure these risk factors. The Treatment Intervention and Progress Scale for Sexual Abusers with Intellectual Disabilities (TIPS-ID) is composed of 25 dynamic risk factors linked to sexual reoffense. Data analyses indicate support for the item composition and reliability of the scale. Ratings from this scale differentiated participants on several clinical variables related to problem severity but not on a measure of static risk. These results are discussed in terms of their clinical and research implications.

Full text

䉷 American Association on Intellectual and Developmental Disabilities 221
VOLUME
112,
NUMBER
3: 221–229 円
MAY
2007
AMERICAN JOURNAL ON MENTAL RETARDATION
A Structured Method of Assessing Dynamic Risk
Factors Among Sexual Abusers With Intellectual
Disabilities
Robert J. McGrath
McGrath Psychological Services (Middlebury, Vermont)
Joy A. Livingston
Flint Springs Associates (Hinesburg, Vermont)
Gail Falk
Vermont Department of Disabilities, Aging and Independent Living (Waterbury)
Abstract
The nature and severity of dynamic risk factors among a group of 87 adult male sexual
abusers with intellectual disabilities were examined as was the psychometric properties of
a new scale designed to measure these risk factors. The Treatment Intervention and Progress
Scale for Sexual Abusers with Intellectual Disabilities (TIPS-ID) is composed of 25 dynamic
risk factors linked to sexual reoffense. Data analyses indicate support for the item com-
position and reliability of the scale. Ratings from this scale differentiated participants on
several clinical variables related to problem severity but not on a measure of static risk.
These results are discussed in terms of their clinical and research implications.
Persons with intellectual disabilities who com-
mit sexual abuse are a small, but increasing con-
cern of the criminal justice and social service sys-
tems (Lindsay, 2002; Lund, 1992). The prevalence
rate of sexual offending among persons with in-
tellectual disabilities appears to be at least as high
as among the nondisabled population (Day, 1994;
Hodgins, 1992; Lindsay, 2004), and treatment
programs for sexual abusers with intellectual dis-
abilities are proliferating (McGrath, Cumming, &
Burchard, 2003). However, a recent descriptive
analysis of 31 studies indicates that treatment eval-
uation efforts with this population are at an early
stage (Courtney & Rose, 2004).
In the absence of a solid research base about
what constitutes effective treatment with this pop-
ulation, service providers for sexual abusers with
intellectual disabilities arguably should follow
principles of intervention that have proven effec-
tive with the general offending population. Two
best practice principles that form the cornerstone
of correctional intervention programs in numer-
ous jurisdictions throughout the world are those
of risk and need (Andrews & Bonta, 2003; Hollin,
2002). The risk principle means that the level of
services provided should correspond to the risk
level of the offender. Those at higher risk are al-
located to higher intensity treatment and super-
vision programs and those with lower risk, to low-
er intensity programs. The need principle means
that treatment interventions should focus on
changeable problems causally linked to the of-
fending behavior, commonly called dynamic risk
factors.
Application of the risk–need model requires
that both an offender’s risk level and treatment
needs be identified, ideally through validated as-
sessment instruments. Some of these instruments

222 䉷 American Association on Intellectual and Developmental Disabilities
VOLUME
112,
NUMBER
3: 221–229 円
MAY
2007
AMERICAN JOURNAL ON MENTAL RETARDATION
Assessing risk factors among sexual abusers R. J. McGrath, J. A. Livingston, and G. Falk
are comprised solely of static risk factors (i.e., un-
changeable historical variables, such as the num-
ber of prior sexual offense convictions); others,
only dynamic risk factors (i.e., potentially change-
able offense-related aspects of an individual’s
functioning, such as pro-offending attitudes), and
still others combine both. Although correctional
researchers have designed most risk and need in-
struments with the general criminal population in
mind, a few investigators have assessed the effi-
cacy of these instruments for use with offenders
who have intellectual disabilities.
In terms of static risk assessment instruments,
the Rapid Risk Assessment for Sexual Offense Re-
cidivism–hereafter called Rapid Risk Assessment
(Hanson, 1997)
is a commonly used actuarial
risk-estimation instrument for adult sexual offend-
ers and has been found to have moderate predic-
tive validity in multiple replication studies (Dor-
en, 2002), one of which was composed entirely of
sexual offenders with intellectual disabilities
(Tough, 2001). More recently, the Violence Risk
Appraisal Guide (Quinsey, Harris, Rice, & Cor-
mier, 2006), an extensively studied actuarial risk
instrument used for predicting violent reoffend-
ing, has been found to have moderate predictive
accuracy among general criminal offenders with
intellectual disabilities (Quinsey, Book, & Skilling,
2004).
Static risk instruments are effective in pre-
dicting the long-term reoffense risk of offenders,
but, because they are composed of unchangeable
risk factors, do not provide direction about how
to reduce that risk. In contrast, dynamic risk in-
struments are useful for identifying key targets for
treatment and supervision, measuring client pro-
gress in these areas and predicting when an indi-
vidual is at increased risk over the short-term. Un-
fortunately, research on dynamic risk instruments
for use with sexual abusers with intellectual dis-
abilities has been limited.
Green, Gray, and Willner (2002) adapted a
dynamic sex offender risk assessment instrument,
the Structured Anchored Clinical Judgment pro-
tocol (Hanson & Thornton, 2000), for use with
men who have intellectual disabilities who had
engaged in inappropriate sexual behavior. Because
this was a preliminary effort, the authors did not
examine the instrument’s interrater reliability,
ability to assess treatment progress, or predictive
validity among this population. In a recent study,
Webster et al. (2006) found that considerable
training and clinical experience may be required
to score this protocol reliably.
A promising dynamic risk instrument used
with the general adult sexual abuser population is
the Sex Offender Need Assessment Rating (Han-
son & Harris, 2001), which is composed of nine
dynamic risk factors and designed for use by pro-
bation and parole officers supervising sexual abus-
ers in the community. In the scale’s development
sample, the total score showed moderate ability to
differentiate between sexual and nonsexual recid-
ivists. Because items selected were limited to those
most highly correlated with sexual recidivism in
the development study, it does not include all of
the factors that potentially are legitimate targets of
treatment in sexual abuser rehabilitation pro-
grams.
Finally, the Sex Offender Treatment Needs
and Progress Scale (McGrath & Cumming, 2003)
was adapted for use in the present study. This
scale is a provider-administered dynamic measure
designed to aid clinicians and probation and pa-
role officers in identifying and monitoring the
treatment needs, supervision needs, and progress
of adult male sex offenders. It is composed of 22
dynamic risk factors empirically or theoretically
linked to sexual offending. Preliminary reports in-
dicate that it can be scored reliably and predicts
sexual reoffending with moderate accuracy
(McGrath, Cumming, & Livingston, 2005).
The lack of established dynamic risk instru-
ments for sexual abusers with intellectual disabil-
ities and concerns that some of the characteristics
and life circumstances of this population were not
addressed by the best existing instruments moti-
vated development of the present scale, the Treat-
ment Intervention and Progress Scale for Sexual
Abusers With Intellectual Disabilities (TIPS-ID).
First, the senior author facilitated work groups in
which developmental services case managers and
treatment staff identified factors that their field
experience had led them to believe were related
positively or negatively to reoffending among sex-
ual abusers with intellectual disabilities. The list
of factors identified by the work groups corre-
sponded closely to factors measured by the Sex
Offender Treatment Needs and Progress Scale
(McGrath & Cumming, 2003). As a result, this
scale was adopted as a framework and modified
to create the current scale.
Three major modifications were made. Be-
cause sexual abusers with intellectual disabilities,
compared to their counterparts with no intellec-

䉷 American Association on Intellectual and Developmental Disabilities 223
VOLUME
112,
NUMBER
3: 221–229 円
MAY
2007
AMERICAN JOURNAL ON MENTAL RETARDATION
Assessing risk factors among sexual abusers R. J. McGrath, J. A. Livingston, and G. Falk
Table 1. Means, SDs, and Endorsement Patterns of Treatment Intervention and Progress Scale for
Sexual Abusers With Intellectual Disabilities (TIPS-ID) Items (N ⫽ 87)
TIPS-ID risk factor Mean SD
Percentage item endorsement
a
0123Missing
1. Admission of offense behavior 1.20 1.06 31.0 33.3 18.4 16.1 01.1
2. Acceptance of responsibility 1.29 1.06 26.4 35.6 18.4 18.4 01.1
3. Sexual behavior 0.85 1.02 47.1 33.3 06.9 12.6 00.0
4. Sexual attitudes 1.51 1.00 17.2 33.3 28.7 19.5 01.1
5. Sexual interests 1.67 1.04 13.8 34.5 23.0 28.7 00.0
6. Sexual knowledge 1.51 0.91 12.6 40.2 31.0 16.1 00.0
7. Criminal and rule-breaking
behavior
1.10 0.95 31.0 36.8 23.0 09.2 00.0
8. Criminal and rule-breaking
attitudes
1.16 0.91 28.7 32.2 33.3 05.7 00.0
9. Substance abuse 0.05 0.26 96.6 02.3 01.1 00.0 00.0
10. Emotion management 1.51 0.82 06.9 49.4 29.9 13.8 00.0
11. Mental health stability 0.91 0.86 36.8 40.2 18.4 04.6 00.0
12. Problem-solving 1.84 0.81 04.6 27.6 47.1 20.7 00.0
13. Impulsivity 1.55 0.85 08.0 43.7 33.3 14.9 00.0
14. Employment/school 1.10 0.95 31.0 36.8 23.0 09.2 00.0
15. Money management 1.64 1.16 24.1 18.4 26.4 31.0 00.0
16. Residence 0.52 0.82 64.4 24.1 06.9 04.6 00.0
17. Social influences: Peers 0.90 0.80 34.5 42.5 19.5 02.3 01.1
18. Social influences: Family 0.86 0.92 44.8 28.7 21.8 04.6 00.0
19. Social involvement 1.66 1.02 16.1 26.4 33.3 24.1 00.0
20. Adult love relationship 2.52 0.97 10.3 03.4 06.9 79.3 00.0
21. Cooperation with treatment 0.92 0.97 42.5 31.0 18.4 08.0 01.1
22. Cooperation with supervision 0.87 0.79 34.5 46.0 14.9 03.4 01.1
23. Risk management knowledge 1.72 0.92 09.2 28.7 35.6 20.7 05.7
24. Risk management application 1.79 1.29 23.0 19.5 05.7 46.0 05.7
25. Stage of change 1.67 1.03 13.8 28.7 26.4 25.3 05.7
Total 31.98 11.78
a
Item endorsement values are: 0 ⫽ minimal or no need for improvement, 1 ⫽ some need for improvement, 2 ⫽
considerable need for improvement, and 3 ⫽ very considerable need for improvement.
tual disabilities, commonly lack accurate infor-
mation about sexuality and sexual boundaries
(Coleman & Haaven, 2001), we added a sex
knowledge item. Second, because many sexual
abusers with intellectual disabilities have few so-
cial contacts outside of the home and family sup-
port can be very important, we added a family
influence item. Finally, a global risk management
item was broken into two items: risk knowledge
and risk application.
The 25 dynamic risk factors listed in Table 1
comprised the final scale. Each risk factor and
scoring criteria were described using a 6-month
recency time frame. Items were scored on a
4-point scale ranging from minimal to no need for
improvement to very considerable need for improve-
ment. Scores were recorded on a scoring sheet and
summed to yield a total score.
In the present study we examined the reli-
ability and validity of this scale as well as score
distribution and item endorsement patterns in the
hope that these data would provide a profile of
the types and severity of treatment needs evident
in this population.

224 䉷 American Association on Intellectual and Developmental Disabilities
VOLUME
112,
NUMBER
3: 221–229 円
MAY
2007
AMERICAN JOURNAL ON MENTAL RETARDATION
Assessing risk factors among sexual abusers R. J. McGrath, J. A. Livingston, and G. Falk
Method
Participants
Participants were 87 male sexual abusers with
intellectual disabilities age 18 or older who were
living in the community and received supervision
and supports funded by the Vermont Department
of Disabilities, Aging and Independent Living be-
tween July 1, 2003, and December 31, 2004. This
department is the single state-funding agency for
people with intellectual disabilities. All partici-
pants lived in the community in their own homes
or supervised residences; Vermont closed its only
institution for individuals with intellectual dis-
abilities in 1993 and began operating a totally
community-based system of services. Data on 10
other individuals who met the above criteria were
not available, and these individuals were not par-
ticipants in this study.
The definition of intellectual disability we used
was the same as the Diagnostic and Statistical Man-
ual of Mental Disorders
DSM-IV-TR criteria for
mental retardation (American Psychiatric Associ-
ation, 2000). The term sexual abuser was defined
as someone who is known to have committed a
sexual misbehavior that, if prosecuted, would con-
stitute a criminal sexual offense in Vermont. The
population included 16 offenders committed to
the custody of Vermont Department of Disabili-
ties, Aging and Independent Living after being
found not competent to stand trial (18%); 19 in-
dividuals on probation or furlough or who had
maxed out of sentence (22%); 15 individuals sub-
stantiated for sexual abuse by child or adult pro-
tective services (17%); and 37 individuals known
to have committed sexual offenses and for whom
no legal finding of guilt was made (42%). Other
participant characteristics are reported in the Re-
sults section.
Measures
Rapid Risk Assessment for Sexual Offense Recid-
ivism. This assessment, developed by Hanson
(1997), is a 4-item actuarial risk measure used to
aid in assessing sexual recidivism risk among con-
victed adult male sex offenders. Rapid Risk As-
sessment items are (a) number of prior charges or
convictions for sexual offenses, (b) age at date of
assessment, (c) whether there were any male vic-
tims, and (d) any unrelated victims. Scores fall
into one of six levels reflecting the probability of
sexual reoffending at 5- and 10-year intervals. Be-
cause sexual abusers with intellectual disabilities
frequently are not criminally charged for known
incidents of offending, adapted scoring criteria
were used (Harris, Phenix, Hanson, & Thornton,
2003, p. 17). Namely, sex offenses substantiated
by formal or informal investigations that resulted
in noteworthy sanctions, such as residential
moves or school suspensions, were counted as
both a sex offense charge and conviction.
Procedure
The senior author provided a one-day training
to staff at the state’s 14 developmental service
agencies and focused on teaching the skills nec-
essary to complete demographic, treatment prog-
ress, and offense coding sheets used in the study
and to score participants on the TIPS-ID. Devel-
opmental services staff submitted to Vermont De-
partment of Disabilities, Aging and Independent
Living demographic and offense data, a treatment
progress rating and at least one TIPS-ID rating for
the 87 study participants. The treatment progress
rating involved evaluating each participant’s over-
all treatment progress for the previous 6 months
as significant, some, or none. Two independent
TIPS-ID and progress ratings were available for 40
(46%) of the 87 study participants. The TIPS-ID
scorers (N ⫽ 39) were primarily case managers.
The first and second authors computed Rapid
Risk Assessment scores from data that service co-
ordinators provided.
Results
Participant Characteristics
The mean age of the sample members was
34.4 years (SD ⫽ 12.5, range ⫽ 18 to 70). Con-
sistent with Vermont’s lack of racial diversity, all
but 5 participants (6%) were White. Based on def-
initions established by the Association for the
Treatment of Sexual Abusers (Gordon et al.,
1998), the sample was composed of 23 rapists
(26%), 14 noncontact sex offenders (15%), 10 in-
cest offenders (12%), and 40 child molesters
(46%). Of the 40 child molesters, 26 had molested
at least one boy (30%) and 14 molested girls only
(16%).
Two thirds of participants (66%) were under
24-hour supervision, and one third (34%) received
less that 24-hour supervision. Most participants
(69%) resided in supervised residences. Thirteen
(15%) were working full- or parttime without sup-
ports, 44% were working full- or parttime with

䉷 American Association on Intellectual and Developmental Disabilities 225
VOLUME
112,
NUMBER
3: 221–229 円
MAY
2007
AMERICAN JOURNAL ON MENTAL RETARDATION
Assessing risk factors among sexual abusers R. J. McGrath, J. A. Livingston, and G. Falk
Table 2. Interrater Reliability (ICC) for Each
Treatment Intervention and Progress Scale for
Sexual Abusers With Intellectual Disabilities
(TIPS-ID) Item and Item Totals (N ⫽ 40)
TIPS-ID risk factor ICC
1
ICC
2
1. Admission of offense behavior .48 .65
2. Acceptance of responsibility .64 .78
3. Sexual behavior .79 .89
4. Sexual attitudes .77 .87
5. Sexual interests .75 .86
6. Sexual knowledge .52 .68
7. Criminal and rule-breaking
behavior
.71 .83
8. Criminal and rule-breaking
attitudes .55 .71
9. Substance abuse .66 .80
10. Emotion management .63 .78
11. Mental health stability .72 .83
12. Problem solving .77 .87
13. Impulsivity .70 .82
14. Employment/school .77 .87
15. Money management .65 .79
16. Residence .74 .85
17. Social influences: Peers .62 .76
18. Social influences: Family .61 .75
19. Social involvement .47 .64
20. Adult love relationship .94 .97
21. Cooperation with treatment .67 .80
22. Cooperation with supervision .63 .78
23. Risk management knowledge .73 .84
24. Risk management application .83 .91
25. Stage of change .69 .82
Total .81 .90
Note. ICC ⫽ interclass correlation for a single rating
(ICC
1
) and for averaged ratings (ICC
2
). For all ICCs, p ⬍
.001.
supports, 6% were students, and about one third
(36%) were neither employed nor in school.
The mean IQ of the sample was 61.9 (SD ⫽
6.4, range ⫽ 45 to 74). At the time of the study,
75% of the sample was receiving treatment de-
signed to address their sexually abusive behavior
and 25% were not. Time in treatment data were
not available. In terms of Rapid Risk Assessment
scores, 35% of participants were categorized as
low risk (score ⫽ 0 or 1); 56%, moderate risk
(score ⫽ 2 or 3); and only 9%, high risk (score ⫽
4 or 5).
Descriptive Statistics
The mean score and SDs for individual TIPS-
ID items and the total score are presented in Table
1. The endorsement pattern for TIPS-ID items in
this table depicts the range of types and severity
of treatment needs evident among participants.
The total mean score of 31.98 (SD ⫽ 11.78) was
almost identical to the median (32.00) and the
modal scores (32.00). Divided into quartiles, 25%
of individuals scored 0 to 24; 28%, 25 to 32; 23%,
33 to 39; and 24%, 40 to 60.
Reliability
For 40 participants in the sample (46%), two
independent ratings were available from pairs of
service providers. Because these pairings were ran-
dom, we used the one-way, random-effects AN-
OVA model intraclass correlation coefficient
(ICC) to compute interrater reliability (Shrout &
Fleiss, 1979). As shown in Table 2, the total scale
ICC for a single rating (ICC
1
) was .81 (95% CI,
.68–.90) and for the average of multiple indepen-
dent ratings (ICC
2
), .90 (95% CI, .81–.95),
F(39, 40) ⫽ 9.75, p ⬍ .001. For single scale items,
data in Table 2 show that 80% of the items had
reliabilities at least .63 for a single rating; all of
the items had reliabilities of .64 or above for av-
eraged ratings.
The scale also showed acceptable internal
consistency. Cronbach’s alpha for the total score
was .91 and the Gutman split-half reliability was
.91. The item-total correlations were .30 or above,
p ⬍ .01, for all but two items, 9: substance abuse,
r ⫽ .12, and 20: adult love relationship, r ⫽ .22,
p ⬍ .05. The standard error of measurement
(SEM) using the total score ICC
1
and ICC
2
were
5.14 and 3.73, respectively, both at the 68% con-
fidence level.
Validity
Because the scale was newly developed, par-
ticipant reoffense data were not available to ex-
amine its predictive validity. Given this circum-
stance, predictive validity might be inferred if in-
dividuals with a large number of static risk factors,
as evidenced by higher Rapid Risk Assessment
scores, were to score higher on the TIPS-ID than
would individuals with a few or no static risk fac-
tors. An ANOVA test did not detect significant
differences between mean TIPS-ID scores of par-
ticipants with low, moderate, and high Rapid Risk

226 䉷 American Association on Intellectual and Developmental Disabilities
VOLUME
112,
NUMBER
3: 221–229 円
MAY
2007
AMERICAN JOURNAL ON MENTAL RETARDATION
Assessing risk factors among sexual abusers R. J. McGrath, J. A. Livingston, and G. Falk
Table 3. Relationship of Treatment Need
Indicators to Mean Treatment Intervention and
Progress Scale for Sexual Abusers With
Intellectual Disabilities (TIPS-ID) Scores
Treatment need
variable n
Total TIPS-ID score
Mean SD
Rapid Risk Assessment
score
Low (0,1) 30 30.87 10.62
Moderate (2,3) 49 31.59 12.74
High (4,5) 8 38.50 8.40
Supervision level
Less than 24-hour 30 27.27* 10.45
24-hour 57 34.48 11.77
Diagnosis of paraphilia
No 55 29.25* 11.76
Yes 32 36.66 10.43
Treatment progress
Significant 21 22.86** 9.17
Some 38 33.79 11.66
None 13 38.00 10.37
*p ⬍ .01. **p⬍ .001.
Assessment scores, F(2, 84) ⫽ 1.40, p ⫽ .25 (see
Table 3). Similarly, the correlation between total
TIPS-ID scores and individual Rapid Risk Assess-
ment scores was not significant, r ⫽ .13, and no
pattern of relationship was found when a scatter
plot of these scores was examined.
Other analyses that might provide support for
the validity of the scale examined the relationship
between mean TIPS-ID scores and three clinically
related variables: level of supervision, a DSM-IV-
TR diagnosis of a paraphilia, and provider ratings
of participant treatment progress. These variables,
albeit subjective, were considered proxies reflect-
ing different levels of treatment need. T tests re-
vealed that TIPS-ID scores were significantly high-
er for individuals under 24-hour supervision as
compared to less than 24-hour supervision, t(85)
⫽ 2.81, p ⬍ .01, and for individuals with a diag-
nosis of paraphilia, t(85) ⫽ 2.95, p ⬍ .01. An
ANOVA test found significant differences in
TIPS-ID scores across individuals with varied lev-
els of treatment progress F(2, 69) ⫽ 9.98, p ⬍ .001
(see Table 3). Post hoc analyses using the Bonfer-
roni correction indicated that individuals with sig-
nificant progress had lower scores than those with
some progress, p ⬍ .001, and those with none, p ⬍
.001. There were no significant differences be-
tween individuals with some progress and those
with none.
Discussion
In the present study we examined the nature
and severity of dynamic risk factors among a
group of sexual abusers with intellectual disabili-
ties and conducted an initial psychometric eval-
uation of a scale designed to measure these risk
factors. The nature of this sample was noteworthy
because it was comprised of the near exhaustive
population of identified sexual abusers with intel-
lectual disabilities in a well-defined geographic
area (the state of Vermont) during a prescribed
time period. Participants were easily identified be-
cause one state agency funds their services.
The TIPS-ID showed good overall interrater
reliability, comparing favorably with other simi-
larly constructed forensic clinical rating scales
(Hare, 2003; Webster et al., 2006; Worling, 2004).
This was significant given that several of the items
were inferential in nature, and many of the scorers
had minimal experience working with persons
who have committed sexual offenses. This sug-
gests that most of the scoring criteria for items are
relatively straightforward and that the scoring in-
structions are adequate.
Some individual items, however, had less
than optimal interrater reliability (Items 1, 6, 8,
and 19). Of these, Item 1 (admission of offense
behavior) and Item 19 (social involvement) have
relatively objective scoring criteria, so we suspect
that problems here were related to scorers having
differing amounts of knowledge about clients’ of-
fending histories and social contacts. Item 6 (sex-
ual knowledge) and Item 8 (criminal and rule-
breaking attitudes) are comparatively more subjec-
tive, indicating that improved scoring instructions
or training efforts may be needed to improve scor-
ing consistency.
Although not a psychometric necessity, pre-
dictors of recidivism on most risk assessment in-
struments are at least mildly intercorrelated with
each other, and this was the case with the TIPS-
ID, with two exceptions. With respect to Item 9
(substance abuse), most participants did not have
alcohol or drug problems and those who did lived
in supervised settings, which limited their access
to these substances. Item 20 (adult love relation-
ship) had low correlation with other items, reflect-

䉷 American Association on Intellectual and Developmental Disabilities 227
VOLUME
112,
NUMBER
3: 221–229 円
MAY
2007
AMERICAN JOURNAL ON MENTAL RETARDATION
Assessing risk factors among sexual abusers R. J. McGrath, J. A. Livingston, and G. Falk
ing the fact that very few participants lived with
or had an adult lover. The lack of correlation for
these two items most likely reflects the limited
social networks of most Vermonters with intellectual
disabilities who receive developmental services.
Endorsement patterns suggested participants
had considerable or very considerable treatment
needs in 11 of the 25 areas assessed (defined by a
mean item score of 1.50 or above). Whether these
and the other factors assessed on the TIPS-ID will
be found to be closely linked to sexual offending
among sexual abusers with intellectual disabilities
remains an empirical question. However, follow-
ing development of the scale, the results of what
is currently the largest meta-analysis of dynamic
risk factors among the general sexual abuser pop-
ulation became available (Hanson & Morton-
Bourgon, 2004, 2005), which supports the rele-
vance of most TIPS-ID items as important treat-
ment targets. Similarly, recently published re-
search by Lindsey, Elliot, and Astell (2004) with
a small sample of sexual abusers who have intel-
lectual disabilities lends support to the scale’s item
composition.
Preliminary findings in the current study pro-
vide some optimism for the validity of the scale.
Participants who were supervised at the most in-
tense level, diagnosed as having a paraphilia, or
judged to have made poor treatment progress had
statistically significant higher TIPS-ID scores than
did those who were not. Participants who had
these characteristics arguably could be expected to
have multiple and elevated dynamic risk factors
and they did indeed, as measured by their TIPS-
ID scores. Although mean TIPS-ID scores were
not statistically significantly related to a measure
of static risk, they were in the expected direction.
Future research on the scale’s predictive validity
should involve follow-up periods of 5 years or
more. This is because the base rate of sexual reof-
fending is quite low over the short-term (Hanson
& Bussier, 1998), and, therefore, it can be difficult
to find statistically significant results when con-
ducting recidivism research with sexual abusers us-
ing short follow-up periods (Barbaree, 1997).
Whether or not total mean TIPS-ID scores are
eventually found to predict sexual reoffending in
follow-up studies, examination of score profiles
may prove valuable. For example, we hypothesize
that individuals who have considerable treatment
needs in the areas of sexual behavior, attitudes,
and interests (Items 3, 4, 5, and 13) will be at
higher risk for sexual reoffending than individuals
who have a similar total mean score but whose
primary problems concern lifestyle instability
(Items 14, 15, and 16). Individuals with high treat-
ment needs in both of these broad areas, sexual
deviancy and lifestyle instability, may be at high-
est risk. Future researchers should include cluster
analytic methods and examine differential weigh-
ing of risk items.
Despite encouraging results in the present
study and other recent research in the field, risk
assessment with this population is quite new. Nev-
ertheless, providers on a daily basis must assess the
likelihood that individuals under their care will
commit new offenses, and program managers
must assess the efficacy of costly treatment inter-
ventions. These are critically important assess-
ments, informing decisions on placement, super-
vision, and treatment. General guidance from the
research literature is at least clear that structured,
evidence-based, risk-assessment approaches are
much more accurate than unaided clinical ap-
proaches (Andrews & Bonta, 2003; Hansen & Bus-
siere, 1998).
Based on this best-practice guidance, we have
some recommendations about how to approach
risk assessment with sexual abusers who have in-
tellectual disabilities. First, assessors should estab-
lish a general risk level using a validated static risk
instrument. This will provide a moderately accu-
rate long-term prediction of the individual’s risk
to sexually reoffend. Currently, the Rapid Risk As-
sessment (Hanson, 1997) appears to be the best
choice (Tough, 2001). Second, assessors should
use a structured method of evaluating the pres-
ence and severity of dynamic risk factors in order
to adjust this risk level, if necessary, and identify
relevant targets of treatment and supervision. For
this purpose, the current scale is a possible mea-
sure. Third, staff members or therapists should pe-
riodically readminister the chosen dynamic instru-
ment to reassess the client’s progress and the ef-
ficacy of treatment interventions and adjust the
risk level, supervision plan, treatment targets, and
treatment focus accordingly. When scores on a
particular item are consistently high for a popu-
lation group (as here for adult love interest), focus
may need to shift from individual treatment to
systemic changes. Finally, evaluators must consid-
er relevant risk factors not commonly accounted
for by structured risk instruments, such as ex-
pressed intent to reoffend or disability that limits
access to potential victims.
The results of this study must be considered

228 䉷 American Association on Intellectual and Developmental Disabilities
VOLUME
112,
NUMBER
3: 221–229 円
MAY
2007
AMERICAN JOURNAL ON MENTAL RETARDATION
Assessing risk factors among sexual abusers R. J. McGrath, J. A. Livingston, and G. Falk
preliminary. Independent use of total scores to
make clinical or legal decisions is not warranted
by the current findings. Providers may most pro-
ductively use the TIPS-ID as a structured method
of periodically examining client progress against a
relatively comprehensive list of clinically and em-
pirically derived risk factors thought to be closely
linked to sexual reoffending. These results, we
hope, will motivate further examination of the in-
strument and increase our knowledge about how
to successfully treat individuals with intellectual
disabilities who have sexually abused others.
References
American Psychiatric Association. (2000). Diag-
nostic and statistical manual of mental disorders
(4th ed., text rev.). Washington, DC: Author.
Andrews, D. A., & Bonta, J. (2003). The psychology
of criminal conduct (3rd ed.). Cincinnati: An-
derson.
Barbaree, H. E. (1997). Evaluating treatment effi-
cacy with sexual offenders: The insensitivity
of recidivism studies to treatment effects. Sex-
ual Abuse: A Journal of Research and Treatment,
9, 111–128.
Coleman, E. M., & Haaven, J. (2001). Assessment
and treatment of intellectually disabled sexual
abusers. In M. S. Carich & S. E. Mussack
(Eds.), Handbook for sexual abuser assessment
and treatment (pp. 193–209). Brandon, VT:
Safer Society.
Courtney, J., & Rose, J. (2004). The effectiveness
of treatment for male sex offenders with learn-
ing disabilities: A review of the literature. Jour-
nal of Sexual Aggression, 10, 215–236.
Day, K. (1994). Male mentally handicapped sex
offenders. British Journal of Psychiatry, 165,
630–639.
Doren, D. M. (2002). Evaluating sex offenders: A
manual for civil commitments and beyond. Thou-
sand Oaks, CA: Sage.
Gordon, A., Harris, A., Murphy, W., Seto, M.,
Hanson, R. K., Marques, J. K., & Quinsey, V.
L. (1998, July). Collaborative data base of sex
offender treatment outcome. Beaverton, OR: As-
sociation for the Treatment of Sexual Abus-
ers.
Green, G., Gray, N. S., & Willner, P. (2002). Fac-
tors associated with criminal convictions for
sexually inappropriate behaviour in men with
learning disabilities. The Journal of Forensic Psy-
chiatry, 13, 578–607.
Hanson, R. K. (1997). The development of a brief
actuarial risk scale for sexual offense recidivism
(User Rep. 97–04). Ottawa: Department of
the Solicitor General of Canada.
Hanson, R. K., & Bussiere, M. T. (1998). Predict-
ing relapse: A meta-analysis of sexual offender
recidivism studies. Journal of Consulting and
Clinical Psychology, 66, 348–362.
Hanson, R. K., & Harris, A. (2001). A structured
approach to evaluating change among sexual
offenders. Sexual Abuse: A Journal of Research
and Treatment, 13, 105–122.
Hanson, R. K., & Morton-Bourgon, K. (2004). Pre-
dictors of sexual recidivism: An updated meta-
analysis. Ottawa: Public Safety and Emergen-
cy Preparedness Canada.
Hanson, R. K., & Morton-Bourgon, K. E. (2005).
The characteristics of persistent sexual offend-
ers: A meta-analysis of recidivism studies.
Journal of Consulting and Clinical Psychology,
73, 1154–1163.
Hanson, R. K., & Thornton, D. (2000) Improving
risk assessments for sex offenders: A compar-
ison of three actuarial scales. Law and Human
Behavior, 24, 119–136.
Hare, R. D. (2003). Hare PCL-R: Technical manual
(2nd ed.). N. Tonawanda, NY: Multi-Health
Systems.
Harris, A., Phenix, A., Hanson, K., & Thornton,
D. (2003). Static-99 coding rules. Ottawa: De-
partment of the Solicitor General of Canada.
Hodgins, S. (1992). Mental disorder, intellectual
deficiency and crime: Evidence from a birth
cohort. Archives of General Psychiatry, 49, 476–
483.
Hollin, C. R. (2002) Risk-need assessment and al-
location to offender programmes. In J.
McGuire (Ed.), Offender rehabilitation and treat-
ment: Effective programmes and policies to reduce
re-offending (pp. 309–332). West Sussex, Eng-
land: Wiley.
Lindsay, W. R. (2002). Research and literature on
sex offenders with intellectual and develop-
mental disabilities. Journal of Intellectual Dis-
ability Research, 46, 74–85.
Lindsay, W. R. (2004). Sex offenders: Conceptu-
alization of the issues, services, treatment and
management. In W. R. Lindsay, J. L. Taylor,
& P. Sturmey (Eds.), Offenders with developmen-
tal disabilities (pp. 163–185). West Sussex,
England: Wiley.
Lindsay, W. R., Elliot, S. F., & Astell, A. (2004).
Predictors of sexual offense recidivism in of-

䉷 American Association on Intellectual and Developmental Disabilities 229
VOLUME
112,
NUMBER
3: 221–229 円
MAY
2007
AMERICAN JOURNAL ON MENTAL RETARDATION
Assessing risk factors among sexual abusers R. J. McGrath, J. A. Livingston, and G. Falk
fenders with intellectual disabilities. Journal of
Applied Research in Intellectual Disabilities, 17,
299–305.
Lund, C. A. (1992). Long-term treatment of sexual
behavior problems of adolescent and adult
developmentally disabled persons. Annals of
Sex Research, 5, 5–31.
McGrath, R. J., & Cumming, G. F. (2003). The sex
offender: Treatment Needs and Progress Scale.
Unpublished test, Middlebury, VT.
McGrath, R. J., Cumming, G. F., & Livingston, J.
A. (2005). Predictive validity of the Sex Offender
Treatment Needs and Progress Scale. Paper pre-
sented at the annual conference of the Asso-
ciation for the Treatment of Sexual Abusers,
Salt Lake City.
McGrath, R. J., Cumming, G. F., & Burchard, B.
L. (2003). Current practices and trends in sexual
abuser management: The Safer Society 2002 Na-
tionwide Survey. Brandon, VT: Safer Society.
Quinsey, V. L., Book, A., & Skilling T. A. (2004).
A follow–up of deinstitutionalized men with
intellectual disabilities and histories of anti-
social behavior. Journal of Applied Research in
Intellectual Disabilities, 17, 243–253.
Quinsey, V. L., Harris, G. T., Rice, M. E., & Cor-
mier, C. A. (2006). Violent offenders: Appraising
and managing risk (2nd ed.). Washington, DC:
American Psychological Association.
Shrout, P. E., & Fleiss, J. L. (1979). Interclass cor-
relations: Uses in assessing rater reliability.
Psychological Bulletin, 86, 420–428.
Tough, S. E. (2001, November). Validation of two
standard risk assessments (RRASOR, 1997; Stat-
ic-99, 1999) on a sample of adult males who are
developmentally disabled with significant cognitive
deficits. Paper presented at the annual confer-
ence of the Association for the Treatment of
Sexual Abusers, San Antonio, TX.
Webster, S. D., Mann, R. E., Carter, A. J., Long,
J., Milner, R. J., O’Brien, M. D., Wakeling,
H. C., & Ray, N. L. (2006). Inter-rater reli-
ability of dynamic risk assessment with sexual
offenders. Psychology, Crime and Law, 12, 439–
452.
Worling, J. R. (2004). The Estimate of Risk of Ad-
olescent Sexual Offense Recidivism (ERA-
SOR): Preliminary psychometric data. Sexual
Abuse: A Journal of Research and Treatment, 16,
235–254.
Received 3/12/06, accepted 11/12/06.
Editor-in-charge: William E. MacLean, Jr.
Research support for this study was provided from
the Office of Justice Programs, U.S. Department
of Justice, Grant No. 20001-WP-BX-0019. The
opinions expressed in this article do not necessar-
ily reflect the views of the Vermont Department
of Disabilities, Aging and Independent Living. Re-
quests for reprints should be sent to Robert J.
McGrath, McGrath Psychological Services, P.C.,
105 Happy Valley Rd., Middlebury, VT 05753.
E-mail: rmcgrath@sover.net