Content uploaded by Rachel L Fazio
Author content
All content in this area was uploaded by Rachel L Fazio on Apr 26, 2014
Content may be subject to copyright.
Open Access Journal of Forensic Psychology
http://www.forensicpsychologyunbound.ws/ -- 2012: 69-80
An Estimate of the Prevalence of Autism-Spectrum Disorders
in an Incarcerated Population
Rachel L. Fazio, Christina A. Pietz, and Robert L. Denney
The School of Professional Psychology at Forest Institute
Springfield, Missouri, USA, Email: rlfazio@gmail.com
Abstract
In the current study, we used the Adult Autism-Spectrum Quotient (AQ) to assess the
prevalence of autism-spectrum disorders (ASD) in a maximum-security prison in the
United States. Results indicated that, employing a conservative cutoff score, 4.4% of
inmates may meet the criteria for a diagnosis of ASD. While this finding is
commensurate with previous research involving other criminal populations, it is four
times greater than the rate of ASD in the general population. The current study is the
first to examine the prevalence of ASD in criminal populations with a U.S. prison
population sample. Given the important implications for both the judicial system and
offender treatment, further research concerning the prevalence of ASD in forensic
populations is needed.
Keywords: Autism, Asperger’s Syndrome, sex offenders
Introduction
Recently, researchers have begun to critically examine the prevalence of individuals
diagnosed with Autism-Spectrum Disorders (ASD) within the criminal-justice system as
well as the implications the diagnosis carries as they progress through legal
proceedings (Browning & Caulfield, 2011). This research adds to the numerous case
studies that have addressed the relation between ASD and violent or sexual offending
behavior (Baron-Cohen, 1988; Chesterman & Rutter, 1993; Haskins & Silva, 2006;
Mawson, Grounds, & Tantam, 1985; Murrie, Warren, Kristiansson, & Dietz, 2002). To
start, Scragg and Shah (1994) examined the prevalence of Asperger’s Syndrome (AS)
in a special hospital in Britain and determined that 1.5% of the population could be
diagnosed with AS. In the event that equivocal cases were included, the prevalence
rose to 2.3%. Similar results were obtained in a later study of special hospitals in
England (Hare, Gould, Mills, & Wing, 1999). In this study, 2.4% of the population of
three special hospitals met criteria for a diagnosis of ASD, which increased to 5.3% with
equivocal or incomplete cases included. Siponmaa, Kristiansson, Jonson, Nyden, &
Gillberg (2001) examined the prevalence of various neuropsychiatric disorders in young
adults referred for pre-sentencing evaluations in Sweden over a five-year period. They
determined none of their patients could be diagnosed with autism, but four percent
could be diagnosed with AS. When probable cases were included, the figure rose to
ten percent. These findings indicate significantly higher rates of ASD among those in
criminal justice settings as compared to the general population. In the United States, for
Autism-Spectrum Disorders
OAJFP – SSN 1948-5115 – Volume 4. 2012
70
example, the prevalence of ASD is currently estimated at .9% (Centers for Disease
Control [CDC], 2009).
While the aforementioned research has indicated a higher prevalence of those with
ASD in various forensic settings, Mourisden, Rich, Isager, and Nedergaard (2008)
conducted a retrospective case-control study of those with ASD. They found that, while
those with autism were much less likely to be convicted of a crime (.9%) as compared to
controls (18.9%), 18.4% of those with AS were convicted of criminal offenses as
opposed to 19.6% of matched controls. The same study found those with AS were
significantly more likely to commit arson, and there was a trend towards increased
likeliness to commit sexual offenses. No differences were found, however, with respect
to other offense types. In another, smaller community-based study, Woodbury-Smith,
Clare, Holland, and Kearns (2006) found individuals with ASD were less likely to offend
compared to a neuro-typical control group, but they were more likely to commit property
damage or acts of violence than controls, who had a higher incidence of drug offenses.
Another potentially important implication of carrying a diagnosis of ASD in criminal
justice proceedings deals specifically with disposition of the case. It has been argued
the current handling of ASD within the criminal justice system is inconsistent and
insufficient to effectively manage individuals with ASD (Browning & Caulfield, 2011).
Due to a lack of knowledge on the part of an attorney, judge, or jury, an offender
diagnosed with an ASD may be viewed more as a cold, remorseless recidivist, rather
than a person with a mental-health diagnosis or neurobiological deficit. And while it
should be assessed on an individual basis, those with severe ASD may also have
cognitive deficits that would impact their competency-related abilities; a diagnosis of
ASD could also be considered when assessing criminal responsibility or as a mitigating
factor during sentencing (Browning & Caulfield, 2011).
Finally, in the event that the legal system diverts individuals with ASD who become
involved in criminal proceedings into mental-health treatment, there is still the question
of finding appropriate, effective services for them. Hare et al. (1999) found those with
ASD were detained two to three years longer than those with other diagnoses in secure
psychiatric settings. In this same vein, studies concerning treatment of those with ASD
or other cognitive limitations have not always shown promising results. One study
began with ten men with intellectual disabilities (six of whom had ASD) and sexual
offending behavior (Murphy, Powell, Guzman, & Hays, 2007). A modified treatment
group was offered on two occasions: two of the men with ASD who completed the
yearlong group on the first occasion also repeated it on the second occasion. Despite
the extensive treatment, one man with ASD continued his sexual-offending behavior
throughout the time he attended both groups, and three men had committed new sexual
offenses at six-month follow-up, all of whom were on the autism spectrum. If indeed
those with ASD are over-represented in the criminal justice system and also detained
longer, it is of utmost importance to develop effective treatments specifically for this
population to reduce criminal behavior and recidivism rates.
Autism-Spectrum Disorders
OAJFP – SSN 1948-5115 – Volume 4. 2012
71
The Current Study
To date, all existing studies concerning the prevalence of ASD in forensic settings have
been conducted overseas; therefore, the generalizability of the prevalence rates to the
United States criminal justice system is questionable. In addition, studies conducted at
secure psychiatric hospitals and with pre-sentencing evaluees are limited in the extent
to which they can generalize to the broader criminal population. The current study
presents data concerning the prevalence of ASD of those incarcerated in a maximum-
security prison in the Midwestern United States. We hope to provide some indication
that criminal behavior may be more common in individuals with ASD than what has
historically been considered (Kohn, Fahum, Ratzoni, & Apter, 1998).
Participants
Approximately 1800 male inmates residing in a single maximum-security state prison in
the Midwest were asked to participate in the survey. Of this number, 431 (24%)
completed the survey. Participant age ranged from 19 to 74, (Mage = 38, SD = 11.8).
Participants had an average of 2 current convictions (SD = 2.3) and 1 prior conviction
(SD = 2). The average current sentence was 15.8 years (SD = 12.2); the average
previous sentence was four years (SD = 4.3). Most participants (71.9%) scored 5 or
higher on the Violence Rating Scale (VRS). A complete description of the VRS follows.
As can be seen in Table 1, however, this score was largely due to the sex offenders in
the sample being almost entirely contact offenders. The non-sexual offenders had a
wider variety of convictions, such as convictions for driving violations and drug sales.
The majority of the participants in this sample (59.9%) had current or prior convictions
for sexual offenses. This was partially due to the sampling method used for data
collection; sex offenders were housed in separate housing units from non-sex offenders
and not all housing units were sampled. Please see Table 1 for demographic data
reported as a single group as well as by offense group.
Materials and Methods
The Adult Autism-Spectrum Quotient (AQ; Baron-Cohen, Wheelwright, Skinner, Martin,
& Clubley, 2001) was administered to assess autism-spectrum traits. The AQ is a 50-
item self-report instrument designed with ten questions in each of five content areas.
The content areas are social skills, attention switching, attention to detail,
communication, and imagination (Baron-Cohen et al., 2001). Some subsequent
analyses have supported a three-factor structure to improve internal consistency (Hurst,
Mitchell, Kimbrel, Kwapil, & Nelson-Gray, 2007). The three factors were best described
as social skills, detail/patterns, and communication. These areas match the “autistic
triad” of symptoms reported in the DSM-IV-TR (American Psychiatric Association,
2000). The initial psychometric assessment of the instrument demonstrated good test-
retest reliability; ratings were also consistent between self-report and parent-report with
self-report producing slightly more conservative scores (Baron-Cohen et al., 2001).
Additionally, the AQ has been validated cross-culturally (Voracek & Dressler, 2006;
Autism-Spectrum Disorders
OAJFP – SSN 1948-5115 – Volume 4. 2012
72
Wakabayashi, Baron-Cohen, Wheelwright, & Tojo, 2006) and with a clinical sample
referred for suspected ASD (Woodbury-Smith, Robinson, Wheelwright, & Baron-Cohen,
2005) among others.
Previous research has shown the AQ is effective as a general population screener for
ASD using a cutoff score of 32 (Baron-Cohen et al., 2001). However, Woodbury-Smith
et al. (2005) determined a cutoff score of 26 was more effective when screening for
ASD in clinical populations. In the current study, results are presented using the
conservative cutoff score of 32, and in a continuous format where possible, because the
optimal cutoff score for criminal populations has not been determined. As the existing
research suggests, ASD may also be related to arson and sexual offending behavior; in
particular, results for those convicted of sexual offenses will be analyzed separately. No
participants had a conviction for arson, thus eliminating the need for analysis on this
factor.
Demographic data (participant age, race/ethnicity, offense history, sentence, and
educational/reading level) were provided by the State Department of Corrections and
coded by the first author. The participants’ most violent offense was also coded using
the Violence Rating Scale (VRS; Young, Justice, & Erdberg, 1999). A rating of “1” on
the Violence Rating Scale indicates a non-violent crime; a rating of “2” indicates
ambiguous violence such as escape or theft; a rating of “3” indicates a property crime.
At ratings of “4” and above, the crimes become more violent, with a “4” indicating threat
of violence. Attacks on persons receive a rating of “5,” crimes involving a loss of life are
rated as a “6,” and the rating of “7” is reserved for extreme violence, such as serial
murder (Young et al., 1999). The purpose of the VRS was to provide the most efficient
manner of quantifying the array of offenses present in the current study’s sample.
The study was announced in the participants’ housing units. Potential participants
assembled in the common area. The researcher read a script explaining the project
while the surveys were distributed. At the conclusion of the testing script, participants
were free to ask questions and agree or decline to participation. All participants signed
informed consent forms, which were approved by both Forest Institute and the
Department of Corrections. Upon receipt of consent, demographic data were obtained
from inmates’ electronic files within the Department of Corrections.
Results
The mean AQ score was 20 (SD = 6.4). Using the cutoff score of 32, 4.4% of the
sample (n = 19) may meet criteria as having an ASD1. Unfortunately, definitive
diagnosis was not possible due to lack of access to protected health information and
independently validated developmental data. Assuming similar psychometric
1 In the initial study, 80% of those with ASD scored in this range as opposed to only 2% of the general
population. In the validation study in a clinical sample, both the sensitivity and specificity at this cutoff
were approximately 75% resulting in a correct classification rate of 76%.
Autism-Spectrum Disorders
OAJFP – SSN 1948-5115 – Volume 4. 2012
73
functioning to the initial and validation studies, however, the majority of these individuals
are likely correctly classified (Baron-Cohen et al., 2001; Woodbury-Smith et al., 2005).
While the distribution of AQ scores was approximately normal, the distribution of a
number of other variables was not, leading to some non-parametric tests being used for
statistical analyses. Please refer to Table 2 for a distribution of Adult Autism-Spectrum
Quotient scores.
Correlations were conducted between AQ and all demographic variables. AQ showed a
marginally significant correlation with age (Spearman’s r = .10, p = .047). In addition,
age was significantly correlated with eduation (Spearman’s r = -.21, p < .001). The
effect of race on AQ was examined by collapsing participants into Caucasian and non-
Caucasian categories in order to have a sufficient sample size in both categories;
differences between groups were then assessed. Independent samples t-test indicated
that the two groups differed significantly on mean AQ score: t(189.8) = 3.32, p = .001,
with the Caucasian participants having a higher score (M = 20.5, SD = 6.6) than the
non-Caucasian participants (M = 18.4, SD = 5.2). A second independent samples t-test
was conducted to determine if convicted sex offenders scored significantly different on
the AQ compared to non-sex offenders. Results revealed higher mean AQ scores for
the sex-offender group compared to non sex offenders: t(419) = -2.1, p = .038 (M =
20.5, SD = 6.9; M = 19.3, SD = 5.4, respectively).
A series of linear regressions was performed to explore the predictive ability of age,
race, education, and sex-offender status to explain AQ score. The first linear regression
explored the effects of the demographic variables (age, race, and education) on AQ.
The Model was significant: F(3, 415) = 9.1, p < .001, explaining 5.5% of the Model’s
variance. A second regression was performed with sex-offender status added. This
Model was also significant: F(4, 414) = 7.4, p < .001, explaining 5.8% of the variance. In
order to determine if sex-offender status was adding predictive utility to the first Model, a
hierarchical linear regression was conducted with the demographic variables in step one
and sex-offender status in step two. The difference between the steps was not
significant: Fchange(1, 414) = 2.3, p = .133. This suggests that the relation between
sex-offender status and higher AQ score is a result of demographic variables, as sex
offenders were more likely to be Caucasian (χ2 (1, n = 431) = 17.02, p < .001), were
significantly older (M = 41.3 and 33.3, respectively, U = 13454.5, z = -6.99, p <.001),
and had a different distribution of eduational levels (χ2 (3, n = 419) = 16.31, p = .001).
Specifically, 26.4% of the non-sex offenders had less than a high school education
compared to 13.1% of the sex offenders. Alternately, while 10.2% of the non-sex
offenders completed college, 19.8% of the sex offenders had done so.
Discussion
As per normed data on the AQ, the inmates in the current study had a higher rate of
autism-spectrum disorders compared to the general population. Even using a
conservative cutoff score of 32, the estimated rate of ASD would be 4.4%, which is four
times more than that of the general population in the United States. These results are
Autism-Spectrum Disorders
OAJFP – SSN 1948-5115 – Volume 4. 2012
74
particularly remarkable given the recent increase in the diagnosis of autism-spectrum
disorders in the US. Per the age of the participants in this study, they are not
demographically a large part of the “autism boom” of the 90’s. This makes their
estimated prevalence of ASD even higher relative to their age cohort.
This finding is strikingly similar to the preliminary research by other authors concerning
the prevalence of ASD in criminal populations. While the three other studies (Hare et
al., 1999; Scragg & Shah, 1994; Siponmaa et al., 2001) regarding the prevalence of
ASD in criminal settings have been performed in different countries, using different
methods of diagnosis, in different criminal settings, and with different age groups, all
have found a prevalence of ASD between 1.5% and 5.3%, or higher with equivocal
cases included. This relatively stable estimate may indicate that those with ASD are,
indeed, overrepresented in criminal settings at the rate of approximately 3-4%, on
average. As previously mentioned, the interaction of those with ASD with the legal
system raises important questions not only on the front end of judicial proceedings,
such as issues relating to competency and criminal responsibility, but also indicates a
need for mental-health professionals to address the specific treatment needs of this
population to reduce criminal behavior and recidivism.
Several demographic variables correlated with AQ. The correlation of AQ score with
education is not unexpected. Those with more autism-spectrum traits are generally
increasingly likely to have either intellectual impairments or behavioral issues that would
interfere with their educational/reading level. The slight correlation with age is
somewhat harder to explain, but one explanation could be the likelihood that older
individuals may be less concerned with social norms and more comfortable pursuing
solitary hobbies or more set in one’s routines compared to their younger counterparts.
Alternately, length of incarceration was not measured in this study, so this tendency for
solitary activity and routine could also be explained by more adaptation to a correctional
lifestyle. Other studies (CDC, 2009) have shown that Caucasian children do have a
higher prevalence rate of ASD (9.9 per 1,000) compared to both African-American (7.2
per 1,000) and Hispanic children (5.9 per 1,000), making the higher AQ score among
Caucasian participants also expected. While AQ did correlate with these variables, the
correlations, although statistically significant, were small, and explained only 5.5% of the
variance in AQ score. Despite the existing case-study literature suggesting otherwise,
sex-offender status accounted for no significant portion of the variance in AQ score
above that explained by demographic variables.
Although the current study is significant in that it helps to further solidify the emerging
research concerning the prevalence of ASD in criminal populations, it does have a
number of limitations. The AQ may benefit from revision in order to achieve maximum
effectiveness with a criminal population, particularly with sex offenders. Two of the
items on the AQ ask about finding it easy to play games involving pretending with
children. These items were frequently skipped by the participants who were sex
offenders, the majority of whom were convicted of child molestation. Additionally, as the
AQ has not yet been validated with criminal or forensic populations, it is impossible to
Autism-Spectrum Disorders
OAJFP – SSN 1948-5115 – Volume 4. 2012
75
be certain what cutoff score yields the most accurate estimate of ASD in this population.
Further, this study did have a relatively low participation rate (24%). This could lead to
bias in the sample. The low participation rate is believed to be largely due to mistrust of
the prison system among the inmates. When some units were surveyed, it was noted
that legally minded inmates would attempt to dissuade others from participating. This
seemed to be due to a fear that the collected information could be used against them in
some way, such as civil commitment. Lack of an incentive also reduced interest among
the potential participants.
The self-selected nature of the participants could also introduce bias into the sample for
other reasons. For example, the selection method could have reduced the estimated
prevalence of ASD in the sample, as those with ASD tend to dislike changes in routine.
Participation in this study was certainly a change in the normal prison routine. Similarly,
the demographic features of this particular sample are not representative of the general
prison population of the US. This sample tended to be older and Caucasian (average
age 38 years and 78% Caucasian) compared to that of the total prison population of the
U.S. (51% under 35 years old; 33% Caucasian; Bureau of Justice Statistics, 2010).
This is partially due to the over-representation of sex offenders in this sample, which
tended to be older and have a high proportion of Caucasian offenders, although the
racial demographics are also representative of the state in which the data were
collected. The different demographic features and the single physical location of the
sample could indicate that results would not generalize to other criminal populations.
The prevalence of ASD in a general prison or jail population has yet to be fully
established. Although research has been performed concerning the prevalence of
these disorders in secure hospitals, young offenders, and now a sample from a
maximum-security prison, these findings are all from very specific subsets of the
general criminal population. Additionally, each of the studies has been performed in a
different country using different assessment methods. Despite these methodological
disparities, all have shown similarly elevated rates of ASD in criminal populations. This
has a number of implications for the criminal justice system as ASD could conceivably
impact competency to stand trial, provide a basis for an insanity defense, or be
accepted as a mitigating factor during sentencing (see Barry-Walsh & Mullen, 2004, and
Browning & Caulfield, 2011, for a discussion of these possibilities). Given the important
implications for both the judicial system and for offender treatment, further research
concerning the prevalence of ASD in forensic populations should be of high priority.
The rapidly increasing prevalence rates of ASD may be foreshadowing a looming issue
for forensic psychologists and the criminal justice system in general.
Received May 28, 2012; revision submitted August 1, 2012
Autism-Spectrum Disorders
OAJFP – SSN 1948-5115 – Volume 4. 2012
76
References
American Psychiatric Association. (2000). Diagnostic and statistical manual of mental
disorders (4th ed., text revision). Washington, DC: Author.
Baron-Cohen, S. (1988). An assessment of violence in a young man with Asperger’s
Syndrome. Journal of Child Psychology & Psychiatry, 29, 351-360.
Baron-Cohen, S., Wheelwright, S., Skinner, R., Martin, J., & Clubley, E. (2001). The
Autism-Spectrum Quotient (AQ): Evidence from Asperger syndrome/high-
functioning autism, males and females, scientists and mathematicians. Journal of
Autism and Developmental Disorders, 31, 5-17.
Barry-Walsh, J. B., & Mullen, P. E. (2004). Forensic aspects of Asperger’s Syndrome.
The Journal of Forensic Psychiatry and Psychology, 15, 96-107.
Browning, A., & Caulfield, L. (2011). The prevalence and treatment of people with
Asperger’s Syndrome in the criminal justice system. Criminology and Criminal
Justice, 11(2), 165-180.
Bureau of Justice Statistics. (2010, June). Prison Inmates at Midyear 2009 – Statistical
Tables (NCJ 230113). Retrieved from
http://bjs.ojp.usdoj.gov/index.cfm?ty=pbdetail&iid=2200
Centers for Disease Control. (2009, December 18). Prevalence of autism spectrum
disorders --- Autism and developmental disabilities monitoring network, United
States, 2006. Morbidity and Mortality Weekly Report, Surveillance Summaries,
58, (SS10), 1-20.
Chesterman, P., & Rutter, S. (1993). Case report: Asperger’s Syndrome and sexual
offending. Journal of Forensic Psychiatry, 4, 555-562.
Hare, D., Gould, J., Mills, R., & Wing, L. (1999). A preliminary study of individuals with
autistic spectrum disorders in three special hospitals in England. London:
National Autistic Society.
Haskins, B., & Silva, A. (2006). Asperger’s disorder and criminal behavior: Forensic-
psychiatric considerations. Journal of the American Academy of Psychiatry and
the Law, 34, 374-384.
Hurst, R. M., Mitchell, J. T., Kimbrel, N. A., Kwapil, T. K., & Nelson-Gray, R. O. (2007).
Examination of the reliability and factor structure of the Autism Spectrum
Quotient (AQ) in a non-clinical sample. Personality and Individual Differences,
43, 1938-1949.
Autism-Spectrum Disorders
OAJFP – SSN 1948-5115 – Volume 4. 2012
77
Kohn, Y., Fahum, T., Ratzoni, G., & Apter, A. (1998). Aggression and sexual offense in
Asperger’s Syndrome. Israel Journal of Psychiatry and Related Sciences, 35(4),
293-299.
Mawson, D., Grounds, A., & Tantam, D. (1985). Violence and Asperger’s Syndrome: A
case study. British Journal of Psychiatry, 147, 566-569.
Mouridsen, A., Rich, B., Isager, T., & Nedergaard, N. (2008). Pervasive developmental
disorders and criminal behaviour: A case control study. International Journal of
Offender Therapy and Comparative Criminology, 52, 196-205.
Murphy, G., Powell., S., Guzman, A., & Hays, S. (2007). Cognitive-behavioural
treatment for men with intellectual disabilities and sexually abusive behaviour: A
pilot study. Journal of Intellectual Disability Research, 51, 902-912.
Murrie, D. C., Warren, J. I., Kristiansson, M., & Dietz, P. E. (2002). Asperger’s
Syndrome in forensic settings. International Journal of Forensic Mental Health, 1,
59-70.
Scragg, P., & Shah, A. (1994). Prevalence of Asperger’s syndrome in a secure hospital.
British Journal of Psychiatry, 165, 679-682.
Siponmaa, L., Kristiansson, M., Jonson, C., Nyden, A., & Gillberg, C. (2001). Juvenile
and young adult mentally disordered offenders: The role of child neuropsychiatric
disorders. The Journal of the American Academy of Psychiatry and the Law, 29,
420-426.
Voracek, M., & Dressler, S. (2006). Lack of correlation between digit ratio (2D:4D) and
Baron-Cohen’s “Reading the Mind in the Eyes” test, empathy, systemizing, and
autism-spectrum quotients in a general population sample. Personality and
Individual Differences, 41, 1481-1491.
Wakabayashi, A., Baron-Cohen, S., Wheelwright, S., & Tojo, Y. (2006). The Autism-
Spectrum Quotient (AQ) in Japan: A cross-cultural comparison. Journal of Autism
and Developmental Disorders, 36, 263-270.
Woodbury-Smith, M. R., Clare, I. C., Holland, A. J., & Kearns, A. (2006). High
functioning autistic spectrum disorders, offending and other law-breaking:
Findings from a community sample. The Journal of Forensic Psychiatry &
Psychology, 17, 108-120.
Woodbury-Smith, M. R., Robinson, J., Wheelwright, S., & Baron-Cohen, S. (2005).
Screening adults for Asperger syndrome using the AQ: A preliminary study of its
diagnostic validity in clinical practice. Journal of Autism and Developmental
Disorders, 35, 331-335.
Autism-Spectrum Disorders
OAJFP – SSN 1948-5115 – Volume 4. 2012
78
Young, M., Justice, J., & Erdberg, P. (1999). Risk factors for violent behavior among
incarcerated male psychiatric patients: A multimethod approach. Assessment,
6(3), 243-258.
Autism-Spectrum Disorders
OAJFP – SSN 1948-5115 – Volume 4. 2012
79
Table 1. Demographic Characteristics – Total and Offense Group – N = 431
Demographic Characteristics
Sex
Offenders
Non-Sex
Offenders
Total
Race/Ethnicity (%)
Caucasian
84.5
67.6
77.7
Non-Caucasian
15.5
32.4
22.3
Age (SD)
41.3 (11.8)
33.27 (9.9)
38 (11.8)
Total Current Convictions (SD)
2.3 (2.0)
2.3 (2.6)
2.3 (2.3)
Total Previous Convictions (SD)
.65 (1.7)
1.2 (2.3)
.86 (2)
Average Current Sentence (years, SD)
16.9 (11.3)
14.3 (13.4)
15.8 (12.2)
Average Past Sentence (years, SD)
4.0 (3.7)
4.0 (4.9)
4 (4.3)
Reading/Education Level (%)
< 6th grade
8.3
14.4
10.7
≥ 6th grade
4.8
12.0
7.6
HS/GED
67.1
63.5
65.5
College Degree
19.8
10.2
16.2
Violence Rating Scale – Highest Offense (%)
Level 1
0
21.4
8.6
Level 2
0
11.0
4.4
Level 3
0.4
18.5
7.7
Level 4
1.9
15.6
7.4
Level 5
96.1
21.4
66.1
Level 6
1.6
12.1
5.8
Note. Those whose educational/reading level could not be determined (2.6% of the sample) were not
included under reading/education level.
Autism-Spectrum Disorders
OAJFP – SSN 1948-5115 – Volume 4. 2012
80
Table 2. Distribution of Adult Autism-Spectrum Quotient Scores
AQ Scores
Frequency
Percent
3
1
.2
4
1
.2
5
1
.2
6
1
.2
7
4
.9
8
2
.5
9
9
2.1
10
8
1.9
11
8
1.9
12
11
2.6
13
12
2.8
14
24
5.6
15
20
4.6
16
25
5.8
17
31
7.2
18
30
7.0
19
28
6.5
20
29
6.7
21
25
5.8
22
21
4.9
23
22
5.1
24
14
3.2
25
18
4.2
26
15
3.5
27
16
3.7
28
9
2.1
29
7
1.6
30
7
1.6
31
13
3.0
32
8
1.9
33
3
.7
34
3
.7
35
2
.5
36
1
.2
37
1
.2
43
1
.2
