Read CJB298280.qxd text version

Criminal Justice and Behavior

http://cjb.sagepub.com Predictors of Rearrest for Rapists and Child Molesters On Probation

Naomi J. Freeman Criminal Justice and Behavior 2007; 34; 752 originally published online May 14, 2007; DOI: 10.1177/0093854806298280 The online version of this article can be found at: http://cjb.sagepub.com/cgi/content/abstract/34/6/752

Published by:

http://www.sagepublications.com

On behalf of:

International Association for Correctional and Forensic Psychology

Additional services and information for Criminal Justice and Behavior can be found at: Email Alerts: http://cjb.sagepub.com/cgi/alerts Subscriptions: http://cjb.sagepub.com/subscriptions Reprints: http://www.sagepub.com/journalsReprints.nav Permissions: http://www.sagepub.com/journalsPermissions.nav Citations http://cjb.sagepub.com/cgi/content/refs/34/6/752

Downloaded from http://cjb.sagepub.com at SAGE Publications on October 31, 2008

PREDICTORS OF REARREST FOR RAPISTS AND CHILD MOLESTERS ON PROBATION

NAOMI J. FREEMAN

State University of New York at Albany

As a result of sex crimes committed against vulnerable populations such as women and youth, American policy makers have increasingly focused on the risk that released sex offenders pose to the public. This study compared sex offender probationers to understand whether predictors of rearrest differ for rapists and child molesters. After a 3-year follow-up, rapists were significantly more likely than child molesters to be rearrested for a nonsexual offense, whereas a trend in the data indicated that child molesters were more likely than rapists to be rearrested for a sexual offense. Results of the study indicated that criminal histories and offender age were robust predictors for both rapists and child molesters. Keywords: sex offenders; probation; recidivism; risk factors

T

he return of convicted sex offenders to local communities has received a great deal of attention in recent years, both in the public forum and in the criminal justice community. National U.S. statistics indicate that as many as 27% of females and 16% of males were sexually abused as children (see Malesky & Keim, 2001) and that one in every eight women will be raped in their lifetime (see D'Amora & Burns-Smith, 1999). As a result of these sexual crimes committed against vulnerable populations such as women and youth, the nation has increasingly focused on the risk that released sex offenders pose to the public. However, little attention has been given to the management of sex offenders who are under probation supervision. With approximately 60% of all convicted sex offenders being supervised in local communities (Greenfeld, 1997; also see Meloy, 2005; Stalans, 2004), empirical research is needed to determine which factors are related to rearrest for sex offenders under probation supervision. Much of the extant literature regarding the relationship between probation supervision and sexual recidivism has analyzed all types of sex offenders together and has not differentiated between rapists and child molesters. However, there is a degree of specialization in sexual offending, in that those offenders who prefer child victims will repeatedly target children and are unlikely to select adults as victims (Hood, Shute, Feilzer, & Wilcox, 2002). Additionally, child molesters and rapists differ in their motivation, their preferred situational aspects of the offense, and the type of victim sought (Hudson & Ward, 1997; Knight, Rosenberg, & Schneider, 1985; Langton & Marshall, 2001). In general, the majority of child molesters are male and have formed some type of relationship with their victim and/or

AUTHOR'S NOTE: Data for this project were furnished by the New York State Division of Criminal Justice Services (DCJS). However, DCJS was not responsible for the methods of statistical analysis or the conclusions reached. Any opinions and suggestions within this article are those of the author alone and not representative of the views of DCJS. The author would like to thank Dr. Greg Pogarsky and Jeffrey C. Sandler, State University of New York at Albany, for their assistance and comments on earlier drafts. The author would also like to thank the anonymous reviewers and the Criminal Justice and Behavior editorial staff for their insight and editorial suggestions. Correspondence concerning this article can be sent electronically to [email protected]

CRIMINAL JUSTICE AND BEHAVIOR, Vol. 34 No. 6, June 2007 752-768 DOI: 10.1177/0093854806298280 © 2007 American Association for Correctional and Forensic Psychology

752

Downloaded from http://cjb.sagepub.com at SAGE Publications on October 31, 2008

Freeman / SEX OFFENDERS

753

the victim's family. As a result of their timid and unassertive nature, child molesters often find comfort in the relationships formed with their victim, as children are perceived as less threatening than adults (Schwartz & Masters, 1985). The offense is most likely sexual in nature, with violence being a rare component of the victimization (Lanyon, 1986; Porter et al., 2000; Revitch & Weiss, 1962). Despite high-profile cases involving violence, statistics indicate that less than 1% of all sexual offenses involve murder (Freeman-Longo, 1996). In comparison, although rapists display some similar characteristics to child molesters (e.g., low self-esteem and feelings of inadequacy), rapists are more versatile offenders who possess traits comparable to general offenders (see Adler, 1984). They tend to be younger at the time of their first arrest, tend to be nonspecialist offenders, and are more likely to be Black (Adler, 1984; Baxter, Marshall, Barbaree, Davidson, & Malcolm, 1984; Smallbone & Milne, 2000). On average, rapists feel little empathy for their victims, hold negative attitudes toward women (Baxter et al., 1984), and condone the use of violence during commission of their sexual offenses (Terry, 2006). Overall, the empirical literature suggests that rapes are not crimes motivated by sexual desire but that they are motivated by anger and are used as a means to gain power and control. Because of these differences between rapists and child molesters, researchers have advocated studying them separately, with offender categorization based on the victim's age and relationship to the offender (see Firestone et al., 1999; Porter et al., 2000). In an attempt to build on previous studies that examined the relationship between probation supervision and sex offender recidivism (e.g., Hepburn & Griffin, 2004; Kruttschnitt, Uggen, & Shelton, 2000; Meloy, 2005), this study compared rapists and child molesters to understand whether risk factors associated with rearrest differ between the two sex offender subgroups.

SEX OFFENDER RECIDIVISM

Risk factors related to sex offender recidivism have been extensively reviewed in the extant literature. Much of what is known regarding these risk factors has been developed from official reports of incarcerated sex offenders. Although official sex offender recidivism data can help in gaining a better understanding of the criminal justice response to sexual crimes and how these responses have impacted public safety, it is important to note the limitations of using official reports. The U.S. Bureau of Justice Statistics reported that between 1992 and 2000, 65% of attempted rapes and 74% of completed sexual assaults were not reported to the police (Rennison, 2002). Sexual victimization is a profoundly personal, harmful incident that often goes unreported. The fact that the majority of sexual assaults are committed by someone who is known to the victim (often an intimate partner or family member) further contributes to the underreporting of these incidents (Bachman, 1998; Fisher, Daigle, Cullen, & Turner, 2003; Kilpatrick, Whalley, & Edmunds, 2000; Terry, 2006). Despite the underreporting of sexual offenses, it is still valuable to review the empirical research that has examined recidivism rates for sex offenders, while recognizing that these statistics are, most likely, underestimates of the true prevalence. Consistently, empirical research has demonstrated that different groups of sex offenders (i.e., rapists, extrafamilial child molesters, and incest offenders) recidivate at different rates (Hanson & Bussière, 1998; Hanson & Morton-Bourgon, 2004). Hanson (2002) found that incest offenders were the least likely to reoffend (8.4%), rapists with adult female victims recidivated at a rate of 17.1%, and child molesters with extrafamilial victims were the most

Downloaded from http://cjb.sagepub.com at SAGE Publications on October 31, 2008

754

CRIMINAL JUSTICE AND BEHAVIOR

likely to reoffend (19.5%; also see Harris & Hanson, 2004). Similar results were obtained by Serin, Mailloux, and Malcolm (2001), who found that 29.4% of rapists, 11.8% of nonfamilial child molesters, and 4.4% of incest offenders recidivated within 7 years of their release. Furthermore, findings from their analysis demonstrated that rapists fared considerably worse than child molesters, recidivating on average 48 months after release compared to 68 months for child molesters. When recidivism data were analyzed by type of offense, 18.9% of rapists were reconvicted for a sexual offense, whereas 22.1% were reconvicted for nonsexual violent behavior after a 4- to 5-year follow-up. On the other hand, 12.7% of child molesters were reconvicted for a sexual offense, and 9.9% were reconvicted for a nonsexual offense (Hanson & Bussière, 1998). Despite differences in recidivism patterns for rapists, extrafamilial child molesters, and incest offenders, some traits related to sexual recidivism are common among the three subgroups. Motiuk and Brown (1996) found that prior sexual offense history, age, and adult drug abuse were significantly related to sexual recidivism. Additionally, offenders who sexually reoffend are more likely to be younger in age, strangers to their victims, single, unemployed, less educated, and to select male victims (Eisenberg, 1997; Hanson & Bussière, 1998; Harris & Hanson, 2004), as well as to physically harm their victims and to have past supervision violations (Dempster & Hart, 2002). Moreover, sex offenders who reoffend are more likely to have a history of violent behavior, to have previous adult convictions, and to have been sexually abused as children (Berliner, Schram, Miller, & Milloy, 1995). Although the low base rates of sexual recidivism and the underreporting of sexual victimizations can make it difficult to accurately assess the risk convicted sex offenders pose to society (Hanson, 2000; Prentky, Lee, Knight, & Cerce, 1997), taken together, it appears that alcohol or drug use, prior sexual offense history, numerous victims in one incident, and antisocial personality traits are related to an increased risk of sexual recidivism (Hanson, Steffy, & Gauthier, 1993; Langstrom, 2002; Langstrom, Sjostedt, & Grann, 2004; Motiuk & Brown, 1996).

SEX OFFENDER PROBATIONERS

Although there has been limited research investigating sex offenders on probation, the research that has been conducted indicates that reoffense patterns for sex offender probationers mirror those of sex offenders sentenced to prison (Hepburn & Griffin, 2004). Offender's age, prior criminal history, drug use, victim's age, and stable employment have all been found to be related to recidivism for sex offender probationers (Kruttschnitt et al., 2000). Additionally, age, race, lack of an intimate partner, and an unstable residence have been associated with probation failure (Meloy, 2005). Research also suggests that offenders who have a history of substance abuse are more likely to violate the conditions of their probation, whereas substance abuse and a prior criminal record are related to the commission of a new offense. Finally, Meloy's (2005) findings suggest that mandatory jail time as part of the conditions of probation was the most robust predictor of sexual recidivism. Surprisingly, Hepburn and Griffin (2004) found that offense characteristics such as the victim's age and relationship to the offender were unrelated to probation success. In fact, Hepburn and Griffin reported few significant static and dynamic risk factors for the likelihood of a technical violation, probation failure, or new criminal offense with their sample of sex offender probationers in Arizona. The only two variables to significantly predict all three outcomes were the offender's marital status and a problem with alcohol or drug use while on probation.

Downloaded from http://cjb.sagepub.com at SAGE Publications on October 31, 2008

Freeman / SEX OFFENDERS

755

Although several studies (e.g., Block, Vane, Barnes, Kassinove, & Motta, 1986; Hepburn & Griffin, 2004; Kruttschnitt et al., 2000; Meloy, 2005) have notably contributed to the knowledge base regarding risk factors related to sex offender probationers' likelihood of recidivism, a paucity of studies have compared sex offenders to discover whether risk factors vary for different types of sex offenders (i.e., rapists and child molesters).

PURPOSE

The majority of research examining sex offender management focuses on sex offenders released from prison and/or the success of prison sex offender treatment programs (Hepburn & Griffin, 2004). Therefore, much of the current literature regarding risk factors is based on sex offenders whose crimes and/or criminal histories are severe enough to require prison terms. As a result, little is known about sex offenders who remain in their communities under probation supervision. Thus, this study attempted to compare the predictors of rearrest for rapists and child molesters on probation. More specifically, the study examined (a) whether there were differences in sexual and nonsexual offense rearrest rates between rapists and child molesters and (b) whether different predictors were related to the commission of subsequent sexual and nonsexual offenses for rapists and child molesters. This research was notable in that it provided a 3-year follow-up, compared rapists and child molesters, and utilized a large sample (Meloy, 2005; Romeo & Williams, 1985).

METHOD

For the purposes of this study, a sex offender was defined as a male convicted of a registerable sexual offense in New York State. Although this population included all sex offenders affected by legislative mandates (i.e., registration and community notification), it did not encompass those offenders who committed a sexual offense and (a) were not convicted or (b) pled to a lesser offense that did not require registration on the state sex offender registry. However, the definition of sex offender included all sex offenders who were known to authorities and were required to register with the state. The data were retrieved from two sources. First, information was obtained from the New York State sex offender registry. New York State, in compliance with federal regulations, passed the Sex Offender Registration Act in 1995, which established the registry on January 21, 1996. It contains information for all registered sex offenders in New York State including offender demographics, offense characteristics, and victim information. At the time of sentencing, the court assigns sex offender probationers a risk level indicating the offender's likelihood of reoffending, as well as his level of dangerousness to the community. Sex offenders in New York State can be classified into three risk levels: Level 1 (low risk), Level 2 (moderate risk), or Level 3 (high risk). Decisions regarding risk levels are made based on the individual facts of each case including, but not limited to, (a) age of the victim(s), (b) number of victims, (c) offender's relationship to the victim(s), (d) duration of the offense conduct with the victim(s), (e) use of weapons or force during commission of the crime, (f) whether the offender was under the influence of alcohol/drugs at the time of the offense, (g) the extent to which the victim(s) was assaulted or injured, and (h) the offender's age at his first sexual misconduct (Division of Criminal Justice Services, 2004).

Downloaded from http://cjb.sagepub.com at SAGE Publications on October 31, 2008

756

CRIMINAL JUSTICE AND BEHAVIOR

Second, criminal history information was extracted for all registered sex offenders from the New York State computerized criminal history database. Criminal history files contain information regarding characteristics related to arrest, conviction, disposition, and sentencing events. As only New York State criminal history information was obtained, crimes that occurred in other states were not included in this study.

PARTICIPANTS

As of August 2005, there were 21,355 sex offenders registered in New York State. To compare predictors related to rearrest for both sexual and nonsexual offenses for rapists and child molesters, only those offenders who were currently under probation supervision were included in the sample (n = 9,170; 42.9%). Therefore, registered sex offenders who were under parole supervision (n = 7,618; 35.7%), those who were supervised by another state (n = 1; < 0.1%), and those who were not under supervision at the time of the study or whose supervising agency was unknown (n = 4,566; 21.4%) were dropped from the study. In addition, it has been well established that female sex offenders are different from male sex offenders (see Vandiver & Kercher, 2004). Therefore, all female sex offenders were dropped from the study (n = 262; 1.2%). Additionally, to ensure an adequate follow-up period for all offenders, only those sex offenders who had been under probation supervision for 3 years were included in the study. Therefore, the final sample was comprised of only male registered sex offenders under probation supervision who had at least a 3-year follow-up period (n = 5,331; 25.0%). A majority of the sex offenders were White (74.4%), whereas 23.5% were Black, and 2.1% were categorized as Indian or Asian. The average registered sex offender was 34.95 years old (SD = 13.09) at the time of his first sexual offense arrest and 35.57 years old (SD = 13.03) at the time he was arrested for the registerable sexual offense, with a range of between 15 and 85 years of age. Most sex offender probationers were registered for sexual contact (n = 2,279; 46.1%) or sexual intercourse (n = 1,693; 34.2%), with the remaining offenders registered for committing deviate sexual intercourse (n = 830; 16.8%), promoting or possessing sexual performance by a child (n = 136; 2.8%), kidnapping or unlawful imprisonment (n = 7; 0.1%), or patronizing/ promoting prostitution (n = 3; 0.1%). Additionally, most sex offender probationers had a criminal history that contained at least one prior arrest (n = 5,185; 97.3%) and 92.5% (n = 4,931) had previously been convicted of at least one felony offense. However, less than 10% had been convicted of more than one felony offense prior to being listed on the sex offender registry. Offenders were followed for 3 years subsequent to the date of their registration on the sex offender registry. The follow-up period was ceased prior to the end of the study if the offender was arrested for a new criminal offense prior to August 2005.

DEPENDENT VARIABLES

As research has indicated differences in rearrests for both sexual and nonsexual offenses among rapists and child molesters (Greenfeld, 1997), and because of the equivocal nature of how to define sex offender recidivism (see Kruttschnitt et al., 2000), two measures of recidivism were used in this study: (a) rearrest for a registerable sexual offense and (b) rearrest for

Downloaded from http://cjb.sagepub.com at SAGE Publications on October 31, 2008

Freeman / SEX OFFENDERS

757

any nonsexual offense. For the purposes of the study, a registerable sexual offense was defined as any sexual crime that resulted in mandated registration with the New York State sex offender registry as stipulated in Corrections Law Section 6c.1 Each rearrest measure was a dichotomous indication of whether the offender was rearrested for the specific type of offense.

INDEPENDENT VARIABLES

Taking into account the risk factors reported by prior research, this study controlled for several factors that have been shown to impact sex offender recidivism, including offender demographics, offender prior criminal history, offense characteristics, and victim information. Table 1 displays frequencies and percentages for the variables, and Table 2 presents the correlation matrix. Offender demographics. To avoid overestimates or underestimates of sexual rearrests (English, Pullen, & Jones, 1997; Hanson & Bussière, 1998; Meloy, 2005; Romeo & Williams, 1985), sex offenders were separated into two distinct subgroups: (a) rapists (n = 631; 11.8%) and (b) child molesters (n = 4,700; 88.2%).2 This variable was computed by analyzing the penal code of the registerable sexual offense as well as the victim's age. Because the majority of sex offenders in the current sample were White, race of the offender was dummy coded as White (n = 3,768; 70.7%; coded as 0) and non-White (n = 1,563; 29.3%; coded as 1). Offender's age at the time they were arrested for the registerable offense was also included in the model. Offender's registry risk level was a continuous variable ranging from 1 to 3, with Level 3 indicating those most at risk for future reoffense. This variable was included to control for the varying levels of supervision that may have resulted from an offender's risk level.3 Finally, the model included dummy variables representing the counties of New York State to control for potential regional impacts and differences in supervision strategies across New York State counties (see Kruttschnitt et al., 2000).4 Prior criminal history. Research has indicated that an offender's prior criminal history is the most robust predictor of future criminal behavior (Abel et al., 1987; Harris & Hanson, 2004; Romeo & Williams, 1985). Therefore, three prior criminal history variables were included in the analyses including: (a) number of prior violent felony arrests, (b) number of prior registerable sexual offense arrests, and (c) number of prior drug offense arrests. In addition to prior arrest variables, number of prior incarceration terms served (both prior jail and/or prison terms)5 and number of prior supervision violations (both prior probation and/or parole supervision violations) were also included as independent variables. Victim information. Several victim variables were included in the analyses, as past research has suggested that victim characteristics are related to offender recidivism (Barbaree & Marshall, 1988; Hanson & Bussière, 1996, 1998). Victim information was based on the instant offense (i.e., the offense that resulted in the offender's mandated registration on the state sex offender registry). The majority of victims were 17 years old or younger (n = 4,281; 84.4%) and female (n = 4,453; 87.3%). In addition to victim age and sex, number of victims in the instant offense was also included, as it has been found to be

Downloaded from http://cjb.sagepub.com at SAGE Publications on October 31, 2008

758

CRIMINAL JUSTICE AND BEHAVIOR

TABLE 1: Registered Sex Offender Demographics

Rapists (n = 631) Offender Demographics

Race White Black Indian or Asian Offender age at instant offense arrest Risk level Level 1 Level 2 Level 3 Prior criminal history Number prior violent felony arrests Number prior registerable sexual offense arrests Number prior drug offense arrests Number prior incarceration terms Number prior supervision violations Victim information Female Male Number of victims 1 2 3 4 5 or more Type of crime for subsequent sex offense arrest Sexual abuse/misconduct Rape Sodomy Kidnapping/unlawful imprisonment Promoting/possessing sexual performance Incest Permitting prostitution Type of crime for subsequent nonsexual arrest Assault or related offense Drug offense Vehicle or traffic law violation Failure to register Larceny Offense related to judicial process Offenses against public order Robbery or related Fraud Firearm Burglary Criminal mischief or related Endangering the welfare of child Other n

Child Molesters (n = 4,700) %

n

%

428 153 19 M = 35.09 M = 1.36 436 161 33

67.83 25.50 3.17 SD = 12.12 SD = 0.58 69.21 25.56 5.24

3,340 1,037 85 M = 35.64 M = 1.59 2,403 1,799 473

71.06 23.24 1.90 SD = 12.92 SD = 0.67 51.40 38.48 10.12

M = 1.22 M = 1.16 M = 0.36 M = 0.57 M = 0.09

545 45 M = 1.08 588 39 3 1 -- 12 5 7 1 -- -- -- 41 32 29 24 15 15 12 10 8 7 7 6 4 11

SD = 1.26 SD = 0.65 SD = 0.98 SD = 1.58 SD = 0.39

93.37 7.63 SD = 0.30 93.19 6.18 0.48 0.16 -- 48.00 20.00 28.00 4.00 -- -- -- 18.55 14.48 13.12 10.86 6.79 6.79 5.43 4.52 3.62 3.17 3.17 2.71 1.81 4.98

M = 0.97 M = 1.20 M = 0.35 M = 0.52 M = 0.08

3,908 605 M = 1.10 4,239 352 34 7 3 119 61 50 26 11 1 1 222 168 156 245 95 95 71 96 42 43 54 46 55 27

SD = 1.12 SD = 0.64 SD = 1.22 SD = 1.67 SD = 0.42

86.59 13.41 SD = 0.36 91.46 7.59 0.73 0.15 0.06 44.24 22.68 18.59 9.67 4.09 0.37 0.37 15.69 11.87 11.02 17.31 6.70 6.70 5.02 6.78 2.97 3.03 3.82 3.25 3.89 1.91

related to offender recidivism (Barbaree & Marshall, 1988; Motiuk & Brown, 1996). Approximately 92% of the offenders had only one victim in the instant offense arrest, with the number of victims ranging from one to nine.

Downloaded from http://cjb.sagepub.com at SAGE Publications on October 31, 2008

Freeman / SEX OFFENDERS TABLE 2: Correlation Matrix

759

1

1. Risk level -- 2. Prior drug arrests .02 3. Prior violent felony arrests .15* 4. Prior sex offense arrests .22* 5. Prior incarceration .08* 6. Prior supervision violations .09* 7. Victim sex .10* 8. Number of victims .15* 9. Offender race ­.04* 10. Offender age at instant .03* offense arrest *p < .01.

2

3

4

5

6

7

8

9

-- .29* .07* .50* .18* ­.05* ­.03* .20* ­.10*

-- .41* .43* .20* ­.05* ­.00 .21* ­.04*

-- .20* .11* .02 .01 .01 ­.00

-- .36* ­.04* ­.02 .12* ­.10*

-- ­.02 ­.02 .03* ­.07*

-- .10* ­.07* .06*

-- ­.06* .10*

-- ­.08*

TABLE 3: Rearrests by Subgroup of Sex Offender

Offense Rearrest Sexual No

Rapists Child molesters Total 606 (96.0) 4,431 (94.3) 5,037 (94.5)

Nonsexual Yes

25 (4.0) 269 (5.7) 294 (5.5)

No

410 (65.0) 3,285 (69.9) 3,695 (69.3)

Yes

221 (35.0) 1,415 (30.1) 1,636 (30.7)

Note. Percentages are given in parentheses. Sexual offense rearrest, 2(1, N = 5,331) = 3.31, p = .07; nonsexual offense rearrest, 2(1, N = 5,331) = 6.32, p = .01.

RESULTS

REARREST RATES

To determine whether there was a significant relationship between type of sex offender and subsequent rearrest for a sexual or nonsexual offense, chi-square tests were conducted. As depicted in Table 3 (and consistent with prior research), the majority of sex offenders were not arrested subsequent to their registration on the sex offender registry. Results of the chi-square test suggested a marginal difference between type of sex offender and subsequent arrest for a sexual offense (p = .07). A trend in the data indicated that child molesters were more likely than rapists to be rearrested for a sexual offense, 5.7% compared to 4.0%. Additionally, results indicated a significant relationship between type of sex offender and subsequent arrest for a nonsexual offense (p = .01), suggesting that rapists were significantly more likely than child molesters to be rearrested for a nonsexual offense, 35.0% compared to 30.1%. A logistic regression was then estimated to examine whether the differences in rearrest rates remained after controlling for risk factors that have been shown to be related to rearrest.6 Results of the aggregate model are presented in Table 4. Consistent with the results of the chi-square test, a marginal difference in rearrest for a sexual offense was found for type of sex offender (p = .08), whereas a significant difference in rearrest for a nonsexual offense emerged for rapists and child molesters (p = .01). Results indicated that the odds of

Downloaded from http://cjb.sagepub.com at SAGE Publications on October 31, 2008

760

CRIMINAL JUSTICE AND BEHAVIOR

TABLE 4: Logistic Regression for Aggregate Model for the Two Outcome Measures

Sexual Offense Rearrest

Child molesters (0) compared to rapists (1) Risk level Number of victims Victim sex (males [0], females [1]) Number of prior drug arrests Number of prior violent felony arrests Number of sexual offense arrests Number of prior incarceration terms Number of prior supervision violations Offender race (White [0], non-White [1]) Offender age at instant offense arrest R2 ­0.53 (.30) ­0.39* (.13) ­0.62* (.23) ­0.51 (.28) ­0.01 (.10) 0.05 (.07) 1.87* (.15) 0.07 (.05) ­0.18 (.18) ­0.28 (.20) ­0.04* (.01) .32

Nonsexual Offense Rearrest

0.25* (.10) 0.13 (.05) ­0.18 (.12) ­0.20 (.11) 0.37* (.09) 0.18* (.05) ­0.63* (.07) 0.33* (.06) ­0.03 (.10) 0.14 (.08) ­0.05* (.00) .14

Note. The model also included fixed-effect county variables. Data are coefficients with robust standard errors in parentheses. *p < .01.

a rapist being rearrested for a sexual offense were 40.9% less than a child molester. In comparison, the odds of a rapist being rearrested for a nonsexual offense were 28.0% higher than a child molester. These results are consistent with prior research (e.g., Adler, 1984; Baxter et al., 1984; Hanson, 2002; Hanson & Bussière, 1998), which has suggested that rapists are significantly more likely than child molesters to be rearrested for a nonsexual offense, whereas child molesters are more likely to sexually reoffend.

PREDICTORS OF REARREST

Given that the results suggest differences in rearrest rates for both sexual and nonsexual offenses for rapists and child molesters, logistic regressions were estimated to determine whether the predictors related to the commission of a subsequent sexual or nonsexual offense varied for rapists and child molesters. Logistic regressions were estimated because the dependent variables of interest were dichotomous. Additionally, logistic regression allows for a determination of the predictors of rearrest for either a sexual or nonsexual offense, while controlling for the other risk factors.7 To provide a means for comparison of the logistic coefficients across models and across the differently measured independent variables, logistic slopes were transformed into predicted probabilities. The transformation of logistic slopes to probabilities, however, requires that a baseline be established. Therefore, the averages of the continuous and dichotomous independent variables (i.e., White offenders and female victims) were selected as the base probability in the transformation calculations. A one-unit change in each independent variable was then calculated, while the other variables remained unchanged. Finally, the base probability was subtracted from the transformed probability (i.e., after the one-unit change). The final transformation yields the probability of rearrest given a one-unit increase in the independent variable, while holding the other predictor variables constant at their means (see Long, 1997). Coefficient comparison tests that were specially designed for logistic regression (Hoetker, 2003) were then computed for those cases in which a predictor variable yielded significant results for at least one of the subgroups of sex offenders. This approach was used rather than ordinary least squares­based coefficient comparison tests (see Paternoster,

Downloaded from http://cjb.sagepub.com at SAGE Publications on October 31, 2008

Freeman / SEX OFFENDERS

761

Brame, Mazerolle, & Piquero, 1998), as the specialized coefficient comparisons allow the residual variation to vary for the two groups of sex offenders. A coefficient comparison test allows one to gauge whether the magnitude of the effects vary between rapists and child molesters. Table 5 summarizes the results of the logistic regression that assessed the predictors of (a) sexual offense rearrest and (b) nonsexual offense rearrest, while controlling for the other factors in the model. Sexual offense rearrest. Consistent with prior research (e.g., Hanson & Bussière, 1996, 1998), results of the analysis indicated that the best predictor of rearrest for a sexual offense for both rapists and child molesters was number of prior sexual offense arrests. Each additional prior sexual offense arrest increased the probability of a rapist being rearrested for a sexual offense by 0.1%, whereas it increased the likelihood of a child molester being rearrested by 15.7%. Number of prior sexual offense arrests was the only variable that was able to significantly predict a subsequent sexual offense arrest for rapists. In contrast, three additional variables emerged as significant predictors for child molesters: (a) victim's sex, (b) number of victims in the instant offense, and (c) offender's age at the time of the instant offense arrest. The probability of child molesters who victimized males being rearrested for a sexual offense was 2.7% greater than the probability of child molesters who selected female victims. Coefficient comparison tests indicated marginally significant differences between child molesters and rapists in terms of victim's sex, indicating that the sex of the victim has a stronger effect on the odds of child molesters, compared to rapists, being arrested for a subsequent sexual offense (p = .07). The number of victims in the instant offense was also a significant predictor of rearrest for a sexual offense for child molesters. Each additional victim in the instant offense increased the probability of a child molester being rearrested for a sexual offense by 2.7%. In other words, child molesters who victimized more than one victim during the instant offense were more likely to be rearrested for a sexual offense than child molesters who had only one victim. However, there was no significant difference in the magnitude of the effect for number of victims between rapists and child molesters (p = .77). Finally, and consistent with prior research, child molesters who were younger at the time of their registerable sexual offense arrest were more likely to be rearrested for a sexual offense. In fact, each 1-year increase in age reduced the likelihood of being rearrested for a sexual offense by 0.3%. Coefficient comparison tests indicated a significant difference in the magnitude of the effect of age for child molesters and rapists (p = .03). No other variables were significant predictors of rearrest for a sexual offense. Nonsexual offense rearrest. A logistic regression was then estimated to compare the predictors of rearrest for nonsexual offenses for rapists and child molesters to discover if these factors varied for each type of sex offender. As depicted in Table 5, four variables emerged as significant predictors for rapists, whereas six variables were able to significantly predict a subsequent nonsexual offense arrest for child molesters. Interestingly, three of the four variables that emerged as significant predictors for rapists (i.e., number of prior drug offense arrests, number of prior sexual offense arrests, and offender's age at the instant offense arrest) were also significant predictors for child molesters. Each additional prior drug offense arrest increased the probability of a rapist being rearrested for a nonsexual offense by 31.5%, whereas it increased the likelihood of a child

Downloaded from http://cjb.sagepub.com at SAGE Publications on October 31, 2008

762

TABLE 5: Logistic Regression for the Comparison Model and the Two Outcome Measures

Sexual Offense Rearrest Rapists One-Unit Changea Coefficient

.01 .03 .03

Nonsexual Offense Rearrest Rapists Coefficient One-Unit Changea Child Molesters Coefficient One-Unit Changea

.03

Child Molesters One-Unit Changea

Coefficient

.32 ­.18 ­.14 ­.02 .27

0.04 (.47) 0.38 (1.30) 1.00 (1.08) ­0.90 (.73) 0.20 (.25) 2.38** (.66) ­0.00 (.12) 1.29 (.88) 0.17 (.67) ­0.01 (.03) .00 .03 ­.00 .33 .46

­0.41** (.14) 0.71** (.23) ­0.62* (.29) 0.03 (.09) 0.04 (.08) 1.93** (.16) 0.05 (.06) ­0.23 (.21) ­0.32 (.12) ­0.05** (.01)

0.08 (.19) ­0.85 (.49) ­0.50 (.43) 1.34** (.25) 0.29 (.16) ­0.87** (.25) 0.25 (.14) ­0.64* (.32) ­0.15 (.25) ­0.06** (.01)

0.14** (.06) ­0.13 (.12) ­0.16 (.11) 0.30** (.09) 0.15** (.05) ­0.62** (.08) 0.35* (.07) 0.02 (.10) 0.18* (.09) ­0.04** (.00) .13

.06 .03 ­.10 .07 .03 ­.01

Downloaded from http://cjb.sagepub.com at SAGE Publications on October 31, 2008

Risk level Number of victims Victim sex (male [0], female [1]) Number of prior drug arrests Number of prior violent felony arrests Number of sexual offense arrests Number of prior incarceration terms Number of prior supervision violations Offender race (White [0], non-White [1]) Offender age at instant offense arrest R2

Note. The model also included fixed-effect county variables. Robust standard errors are presented in parentheses. a. This value represents the predictive probability after a one-unit change in the independent variable. Changes were computed with other variables held at their means (Long, 1997). A one-unit change is only presented for those variables that were significant predictors of the outcome measure. *p < .05. **p < .01, two-tailed.

Freeman / SEX OFFENDERS

763

molester being arrested by 5.7%. Coefficient comparison tests yielded significant differences between type of sex offender and number of prior drug offense arrests, indicating a significant difference in the magnitude of the effect for each type of sex offender (p < .001). Contrary to the results for subsequent sexual offense arrest, number of prior sexual offense arrests actually reduced the probability of being rearrested for a nonsexual offense for both rapists (18.4%) and child molesters (9.1%), suggesting that those sex offenders with lengthy prior sexual offense histories are unlikely to commit nonsexual offenses. This finding is supported by prior work that suggests differences in offending patterns for child molesters (who are more likely to sexually reoffend) compared to rapists (who are more likely to nonsexually reoffend). Although number of prior sexual offense arrests was a significant predictor for both types of sex offenders, coefficient comparison tests indicated no significant differences in its predictive ability for rapists and child molesters (p = .65). Finally, offender's age at the time of the registerable sex offense was a significant predictor for both rapists and child molesters. Each 1-year increase in age reduced the likelihood of a rapist being rearrested for a nonsexual offense by 1.8%, whereas it reduced the probability of a child molester being rearrested for a nonsexual offense by 0.8%. However, when the coefficients for age at the time of the registerable sex offense were compared, no significant differences by type of sex offender emerged (p = .24). Number of prior supervision violations was a significant predictor of subsequent nonsexual offense arrests for rapists, but not for child molesters. Each additional supervision violation decreased the probability of being rearrested for a nonsexual offense by 14.1%. Although the negative relationship may appear counterintuitive, those with prior supervision violations may be supervised more intensively than their counterparts and, therefore, have less opportunities to reoffend. Coefficient comparison tests for prior supervision violations indicated no significant differences by type of sex offender (p = .06). Three additional variables emerged as significant predictors for child molesters, but not for rapists. Each additional incarceration term increased the chance of a child molester being rearrested for a nonsexual offense by 6.7%. Examination of the coefficient comparison test indicated that the magnitude of the effect of prior incarceration terms was significantly different for rapists and child molesters (p < .001). Child molesters with prior violent felony offense arrests were 2.9% more likely to be rearrested for a nonsexual offense than their counterparts. Additionally, the probability of non-White child molesters being rearrested for a nonsexual offense was 3.2% higher than White child molesters. Coefficient comparison tests revealed no significant differences in the magnitude of the effects between rapists and child molesters on either prior violent felony offense arrests (p = .80) or race (p = .16).

DISCUSSION

This study compared sex offender probationers to understand (a) whether rearrest rates differed for rapists and child molesters and (b) whether the predictors of rearrest for sexual and nonsexual offenses varied for rapists and child molesters. With the large percentage of sex offenders under probation supervision and the limited research in this area, the identification of predictors of rearrest can assist in targeting interventions and services to appropriate sex offenders, ultimately increasing the effectiveness of these services and reducing the likelihood of recidivism.

Downloaded from http://cjb.sagepub.com at SAGE Publications on October 31, 2008

764

CRIMINAL JUSTICE AND BEHAVIOR

Before reviewing the findings, it is important to note that the results of this study should not be assumed to generalize to all groups of sex offenders. This study was limited in that the sample included only those male sex offenders who were convicted of a registerable sexual offense and mandated to register with the state sex offender registry. Therefore, the results of this study may not be generalizable to sex offenders whose crimes and/or criminal histories allowed them to plead to lesser nonsexual offenses. Furthermore, this study relied on archival criminal history information that was limited to New York State arrests, which may have contributed to underestimating the frequency of rearrest by convicted sex offenders. Lizotte (1985) suggested that rapes and sexual assaults are more likely to be reported when they have qualities that make them stronger to prosecute. Likewise, an examination of the reasons why these incidents go unreported reveal that female students are more likely to report their sexual victimizations when the assaults have characteristics that make the incidents more believable, such as the use of a weapon or injury, or were perpetrated by a stranger (Fisher et al., 2003). Although variables such as offender risk level (which is composed of weapon use, injury to the victim, and offender-victim relationship) were included in the analyses to control for these effects, it is possible that these factors also contributed to underestimation of sexual rearrests in this study. Examination of the results for the differences in rearrest rates for sexual and nonsexual offenses for rapists and child molesters revealed low rates of rearrests for both types of sex offenders during the 3-year follow-up period, supporting prior research that has also found low rearrest rates for sex offenders (e.g., Harris & Hanson, 2004; Hughes & Wilson, 2003; Meloy, 2005). Of the 5,331 probationers, 30.7% were rearrested for a nonsexual offense after their registration date, with the most common arrests for both rapists and child molesters being assault, drug offenses, vehicle/traffic violations, and failure to register with the sex offender registry. Of the 5.5% of offenders who were rearrested for a sexual offense, most were rearrested for sexual misconduct or abuse, rape, or sodomy (see Table 1 for a complete list of subsequent offense types and corresponding frequencies and percentages). These results are consistent with the findings of Hepburn and Griffin (2004), who found the most common new criminal offenses to be failure to register as a sex offender and violent nonsexual offenses. The most salient difference to emerge between rapists and child molesters was the rearrest rates for the two outcome measures. Results indicated that rapists were more likely than child molesters to be rearrested for a nonsexual offense (rapists 35.0%, child molesters 30.1%), whereas child molesters were more likely to be rearrested for a sexual offense (rapists 4.0%, child molesters 5.7%). These results remained even after controlling for several risk factors and initial differences between the two groups. Research has suggested different offense pathways to recidivism for different types of sex offenders (see Hudson, Ward, & McCormack, 1999; Roberts, Doren, & Thornton, 2002). For example, Knight (1999, as cited in Roberts et al., 2002) argued that there were two pathways to sexual coercion: (a) sexual deviance; and (b) aggression, hostility, and violence toward women. Results of this study support the theory of distinct offense pathways, as does the extant literature, which maintains differences in criminal motivation among rapists and child molesters. Unlike rapists, who are usually motivated by violence, child molesters appear to be motivated by the sexual components of the offense (Porter et al., 2000). In fact, typologies utilized to classify child molesters support the conclusion that most child molesters experience high levels of fixation on children and display low levels

Downloaded from http://cjb.sagepub.com at SAGE Publications on October 31, 2008

Freeman / SEX OFFENDERS

765

of social competence (see Knight, Carter, & Prentky, 1989). Research on rapists, on the other hand, has suggested that rapists display negative attitudes toward women and sex (Baxter et al., 1984), often lack empathy for their victims, and condone violence as a component of the sexual victimization (Terry, 2006). These findings may help explain why rapists in this sample were more likely to be rearrested for nonsexual offenses, whereas child molesters were more likely to be rearrested for sexual offenses. That is, rapes appear to be crimes of opportunity motivated by violence and/or anger. Thus, rapists are versatile offenders, committing both rapes and other general offenses when opportunities arise. On the other hand, child molestations are crimes motivated by sexual desires and deviances. Therefore, child molesters are specialized offenders who rarely commit nonsexual offenses, as these types of offenses do not fulfill their sexual desires. The differences in rearrest patterns for rapists and child molesters yield implications for interventions and services offered to sex offenders. Andrews, Bonta, and Hoge (1990) noted three principles of effective correctional interventions: risk, need, and responsivity. More specifically, Andrews and colleagues argued that interventions and services are most effective when matched with specific offender needs. The importance of targeting programs to specific offender needs was also outlined by Andrews and Kiessling (1980), who found that directing interventions toward an inappropriate level offender resulted in increased recidivism. Given the results of this study that indicate differences in rearrest offense patterns for rapists and child molesters, perhaps specialized treatment programs should be established for each subgroup of sex offender, with child molester programs focusing on sexual deviances and programs for rapists focusing on violence and low self-control. Targeting intervention programs to specific types of sex offenders will more effectively reduce the likelihood of recidivism, ultimately increasing public safety. The second aim of the study was to examine whether predictors of rearrest for sexual and nonsexual offenses varied for rapists and child molesters. Traditional sex offender research maintains that child molesters and rapists have many traits in common, yet they differ on their motivation, their preferred situational aspects of the offense, and the type of victim sought (Hood et al., 2002). Because of these differences, the most widely used sex offender classification scheme involves dichotomizing sex offenders based on victim's age into (a) child molesters and (b) rapists (see Porter et al., 2000). Despite the extant research indicating these differences, this study revealed limited significant differences in the predictors of rearrest for sexual and nonsexual offenses for rapists and child molesters. That is, for both rapists and child molesters, and for both outcome measures (i.e., sexual offense rearrest and nonsexual offense rearrest), criminal history variables and offender's age emerged as robust predictors. These results support prior work that indicates offenders who sexually reoffend are more likely to have a history of sexual and criminal offenses (Dempster & Hart, 2002; Hanson & Bussière, 1998; Hanson, Scott, & Steffy, 1995; Harris & Hanson, 2004; Motiuk & Brown, 1996). The limited differences in the predictors of sexual and nonsexual offense rearrest for rapists and child molesters may indicate that specific predictors of rearrest are not as important as the offense patterns of rapists compared to child molesters. It appears that the predictors of rearrest for a sexual or nonsexual offense vary little between rapists and child molesters, whereas the types of offenses committed distinguish the two groups. Therefore, risk assessment instruments that predict the level of risk for sex offenders and their dangerousness to communities are adequate for both types of sex offenders (i.e., rapists and child molesters).

Downloaded from http://cjb.sagepub.com at SAGE Publications on October 31, 2008

766

CRIMINAL JUSTICE AND BEHAVIOR

Taken together, the results suggest that rapists who are young and have extensive criminal histories are likely to be rearrested for nonsexual offenses, whereas child molesters who are young and have long offense histories are likely to be rearrested for sexual offenses. To maximize the scarce resources that are available for sex offender management, young rapists with diverse criminal histories should be given intensive treatments that address substance abuse, low self-control, and violent tendencies. Young child molesters with long criminal histories, on the other hand, should receive therapy services that address their fixation on children and their sexual deviances. As recommended by Andrews and colleagues (1990), and to lower recidivism rates, higher levels of services should be reserved for higher risk offenders (i.e., offenders who are very young or have a long history of criminal activity). With the trend in the United States toward lifetime probation and civil commitment, predictors of rearrest can be used to target these higher, more intensive interventions toward the appropriate high-risk sex offenders.

NOTES

1. Rearrest was selected as the measure of recidivism, as objections have been raised regarding the use of reconviction. Because sexual crimes are less likely to be reported to authorities and many sexual assault cases are never prosecuted, using reconviction rates may considerably underestimate the rates of recidivism (see Romeo & Williams, 1985). However, it should be noted that an offender's being rearrested for a sexual offense does not signify that he was convicted of that crime, and therefore, it is possible that rearrest data produce false positive results (Romeo & Williams, 1985). 2. Incest offenders and extrafamilial child molesters were grouped together in light of the research that suggests a large proportion of child molesters admit to victimizing both relative and nonrelative children (Heil, Ahlmeyer, & Simons, 2003). 3. Although risk level was composed of other variables in the model, auxiliary regression equations indicated that only 7% of the variation in risk level was shared with the other variables in the model. 4. Separate regression analyses were estimated to determine whether there were extensive differences in the effects of the predictor variables across counties. Counties where less than 40 sex offenders resided (n = 24; 0.5% of the population) were excluded from these tests. Limited significant differences were found across counties. Therefore, the following analyses are based on the aggregate sample. 5. As a result of the correlation between number of prior drug offense arrests and number of prior incarceration terms (r = .50), two separate logistic regression models were estimated eliminating one of the variables from each of the models. Results remained the same with the absence of one of these variables in the models, suggesting no concerns with collinearity or shared variation between the variables. 6. The degree of multicollinearity among the independent variables was assessed by estimating auxiliary regression equations (i.e., additional analyses replacing each independent variable as the dependent variable). When this method is employed, one wants an R2 value less than 75%. Results of the analyses indicated no collinearity among the variables. In addition to assessing multicollinearity by replacing each independent variable as the dependent variable, variance inflation factor coefficients (the proportion of variance of the slope inflated by collinearity) were assessed, also indicating no concerns with multicollinearity. 7. To correct for potential dependency of observations within geographic observations, robust standard errors were estimated to ensure conservative standard errors.

REFERENCES

Abel, G. C., Becker, J. V., Mittelman, M., Cunningham-Rathner, J., Rouleau, J. L., & Murphy, W. D. (1987). Self-reported sex crimes of nonincarcerated paraphiliacs. Journal of Interpersonal Violence, 2, 3-25. Adler, C. (1984). The convicted rapist: A sexual or a violent offender? Criminal Justice and Behavior, 11, 157-177. Andrews, D. A., Bonta, J., & Hoge, R. D. (1990). Classification for effective rehabilitation: Rediscovering psychology. Criminal Justice and Behavior, 17, 19-52. Andrews, D. A., & Kiessling, J. J. (1980). Program structure and effective correctional practices: A summary of the CaVIC research. In R. R. Ross & P. Gendreau (Eds.), Effective correctional treatment. Toronto, Canada: Butterworths. Bachman, R. (1998). The factors related to rape reporting behavior and arrest: New evidence from the National Crime Victimization Survey. Criminal Justice and Behavior, 25, 8-29.

Downloaded from http://cjb.sagepub.com at SAGE Publications on October 31, 2008

Freeman / SEX OFFENDERS

767

Barbaree, H. E., & Marshall, W. L. (1988). Deviant sexual arousal, offense history, and demographic variables as predictors of reoffense among child molesters. Behavioral Sciences and the Law, 6, 267-280. Baxter, D. J., Marshall, W. L., Barbaree, H. E., Davidson, P. R., & Malcolm, P. B. (1984). Deviant sexual behavior: Differentiating sex offenders by criminal and personal history, psychometric measures, and sexual response. Criminal Justice and Behavior, 11, 477-501. Berliner, L., Schram, D., Miller, L. L., & Milloy, C. D. (1995). A sentencing alternative for sex offenders: A study of decision making and recidivism. Journal of Interpersonal Violence, 10, 487-502. Block, J. R., Vane, J., Barnes, M., Kassinove, H., & Motta, R. (1986). A study of sex offenders on probation. Hempstead, NY: Hofstra University Press. D'Amora, D., & Burns-Smith, G. (1999). Partnering in response to sexual violence: How offender treatment and victim advocacy can work together in response to sexual violence. Sexual Abuse: A Journal of Research and Treatment, 11, 293-304. Dempster, R. J., & Hart, S. D. (2002). The relative utility of fixed and variable risk factors in discriminating sexual recidivists and nonrecidivists. Sexual Abuse: A Journal of Research and Treatment, 14, 121-138. Division of Criminal Justice Services. (2004). Sex offender risk level determinations. Retrieved June 28, 2004, from http://criminaljustice.state.ny.us Eisenberg, M. (1997). Recidivism of sex offenders: Factors to consider in release decisions. Austin, TX: Criminal Justice Policy Council. English, K., Pullen, S., & Jones, L. (1997). Managing adult sex offenders in the community--A containment approach. Washington, DC: U.S. Department of Justice. Firestone, P., Bradford, J. M., McCoy, M., Greenberg, D. M., Larose, M. R., & Curry, S. (1999). Prediction of recidivism in incest offenders. Journal of Interpersonal Violence, 14, 511-531. Fisher, B. S., Daigle, L. E., Cullen, F. T., & Turner, M. G. (2003). Reporting of sexual victimization to the police and others: Results from a national-level study of college women. Criminal Justice and Behavior, 30, 6-38. Freeman-Longo, R. E. (1996). Prevention or problem. Sexual Abuse: A Journal of Research and Treatment, 4, 91-100. Greenfeld, L. (1997). Sex offenses and offenders: An analysis of data on rape and sexual assault. Washington, DC: Bureau of Justice Statistics. Hanson, R. K. (2000). Will they do it again? Predicting sex-offense recidivism. Current Directions in Psychological Science, 9, 106-109. Hanson, R. K. (2002). Recidivism and age: Follow-up data from 4,673 sexual offenders. Journal of Interpersonal Violence, 17, 1046-1062. Hanson, R. K., & Bussière, M. T. (1996). Predictors of sexual offender recidivism: A meta-analysis. Ottawa: Solicitor General of Canada. Hanson, R. K., & Bussière, M. T. (1998). Predicting relapse: A meta-analysis of sexual offender recidivism studies. Journal of Consulting and Clinical Psychology, 66, 348-362. Hanson, R. K., & Morton-Bourgon, K. (2004). Predictors of sexual recidivism: An updated meta-analysis. Ottawa: Public Safety and Emergency Preparedness Canada. Hanson, R. K., Scott, H., & Steffy, R. A. (1995). A comparison of child molester and nonsexual criminals: Risk predictors and long-term recidivism. Journal of Research in Crime and Delinquency, 32, 325-337. Hanson, R. K., Steffy, R. A., & Gauthier, R. (1993). Long-term recidivism of child molesters. Journal of Consulting and Clinical Psychology, 61, 646-652. Harris, A.J.R., & Hanson, R. K. (2004). Sex offender recidivism: A simple question. Ottawa: Solicitor General of Canada. Heil, P., Ahlmeyer, S., & Simons, D. (2003). Crossover sexual offenses. Sexual Abuse: A Journal of Research and Treatment, 15, 221-236. Hepburn, J. R., & Griffin, M. L. (2004). An analysis of risk factors contributing to the recidivism of sex offenders on probation. Washington, DC: U.S. Department of Justice. Hoetker, G. (2003). Confounded coefficients: Accurately comparing logit and probit coefficients across groups. Retrieved December 8, 2005, from http://www.business.uiuc.edu/Working Papers/ Hood, R., Shute, S., Feilzer, M., & Wilcox, A. (2002). Sex offenders emerging from long-term imprisonment. British Journal of Criminology, 42, 371-394. Hudson, S. M., & Ward, T. (1997). Rape: Psychopathology and theory. In D. R. Laws & W. O'Donohue (Eds.), Sexual deviance: Theory, assessment, and treatment (pp. 332-355). New York: Guilford. Hudson, S. M., Ward, T., & McCormack, J. C. (1999). Offense pathways in sexual offenders. Journal of Interpersonal Violence, 14, 779-798. Hughes, T., & Wilson, D. J. (2003). Reentry trends in the United States. Washington, DC: Bureau of Justice Statistics, U.S. Department of Justice. Kilpatrick, D. G., Whalley, A., & Edmunds, C. (2000). Chapter 10 sexual assault. In A. Seymour, M. Murray, J. Sigmon, M. Hook, C. Edmunds, M. Gaboury, & G. Coleman (Eds.), 2000 National victim assistance academy. Retrieved April 19, 2005, from http://www.ojp.usdoj.gov/ovc/assits/nvaa2000/academy/J-10-SA.htm Knight, R. A., Carter, D. L., & Prentky, R. A. (1989). A system for the classification of child molesters: Reliability and application. Journal of Interpersonal Violence, 4, 3-23.

Downloaded from http://cjb.sagepub.com at SAGE Publications on October 31, 2008

768

CRIMINAL JUSTICE AND BEHAVIOR

Knight, R. A., Rosenberg, R., & Schneider, B. A. (1985). Classification of sexual offenders: Perspectives, methods, and validation. In A. W. Burgess (Ed.), Rape and sexual assault: A research handbook (pp. 222-293). New York: Garland. Kruttschnitt, C., Uggen, C., & Shelton, K. (2000). Predictors of desistance among sex offenders: The interaction of formal and informal social controls. Justice Quarterly, 17, 61-87. Langstrom, N. (2002). Long-term follow-up of criminal recidivism in young sex offenders: Temporal patterns and risk factors. Psychology, Crime & Law, 8, 41-58. Langstrom, N., Sjostedt, G., & Grann, M. (2004). Psychiatric disorders and recidivism in sexual offenders. Sexual Abuse: A Journal of Research and Treatment, 16, 139-150. Langton, C. M., & Marshall, W. L. (2001). Cognition in rapists: Theoretical patterns by typological breakdown. Aggression and Violent Behavior, 6, 499-518. Lanyon, R. I. (1986). Theory and treatment in child molestation. Journal of Consulting and Clinical Psychology, 54, 176-182. Lizotte, A. J. (1985). The uniqueness of rape: Reporting assaultive violence to the police. Crime & Delinquency, 31, 169-190. Long, J. S. (1997). Regression models for categorical and limited dependent variables. Thousand Oaks, CA: Sage. Malesky, A., & Keim, J. (2001). Mental health professionals' perspectives on sex offender registry Web sites. Sexual Abuse: A Journal of Research and Treatment, 13, 53-63. Meloy, M. L. (2005). The sex offender next door: An analysis of recidivism, risk factors, and deterrence of sex offenders on probation. Criminal Justice Policy Review, 16, 211-236. Motiuk, L. L., & Brown, S. L. (1996). Factors related to recidivism among released federal sex offenders. Ottawa: Research Division Correctional Service Canada. Paternoster, R., Brame, R., Mazerolle, P., & Piquero, A. (1998). Using the correct statistical test for the equality of regression coefficients. Criminology, 36, 859-866. Porter, S., Fairweather, D., Drugge, J., Herve, H., Birt, A., & Boer, D. P. (2000). Profiles of psychopathy in incarcerated sexual offenders. Criminal Justice and Behavior, 27, 216-233. Prentky, R. A., Lee, A.F.S., Knight, R. A., & Cerce, D. (1997). Recidivism rates among child molesters and rapists: A methodological analysis. Law and Human Behavior, 21, 635-659. Rennison, C. M. (2002). Rape and sexual assault: Reporting to police and medical attention, 1992-2000. Washington, DC: Bureau of Justice Statistics, U.S. Department of Justice. Revitch, E., & Weiss, R. (1962). The pedophiliac offender. Diseases of the Nervous System, 23, 73-78. Roberts, C. F., Doren, D. M., & Thornton, D. (2002). Dimensions associated with assessment of sex offender recidivism risk. Criminal Justice and Behavior, 29, 569-589. Romeo, J. J., & Williams, L. M. (1985). Recidivism among convicted sex offenders: A 10-year follow-up study. Federal Probation, 49, 58-64. Schwartz, M. F., & Masters, W. H. (1985). Treatment of paraphiliacs, pedophilias, and incest families. In A. W. Burgess (Ed.), Rape and sexual assault: A research handbook (pp. 350-364). New York: Garland. Serin, R. C., Mailloux, D. L., & Malcolm, P. B. (2001). Psychopathy, deviant sexual arousal, and recidivism among sexual offenders. Journal of Interpersonal Violence, 16, 234-246. Smallbone, S. W., & Milne, L. (2000). Associations between trait anger and aggression used in the commission of sexual offenses. International Journal of Offender Therapy and Comparative Criminology, 44, 606-617. Stalans, L. J. (2004). Adult sex offenders on community supervision: A review of recent assessment strategies and treatment. Criminal Justice and Behavior, 31, 564-608. Terry, K. J. (2006). Sexual offenses and offenders: Theory, practice, and policy. Belmont, CA: Thomson Wadsworth. Vandiver, D. M., & Kercher, G. (2004). Offender and victim characteristics of registered female sexual offenders in Texas: A proposed typology of female sexual offenders. Sexual Abuse: A Journal of Research and Treatment, 16, 121-137.

Downloaded from http://cjb.sagepub.com at SAGE Publications on October 31, 2008

Information

CJB298280.qxd

18 pages

Find more like this

Report File (DMCA)

Our content is added by our users. We aim to remove reported files within 1 working day. Please use this link to notify us:

Report this file as copyright or inappropriate

877424

You might also be interested in

BETA
75301.lay
Microsoft Word - Static-2002R Revised Age Weights _January 2010_.doc
The Psychology of Interviewing, Interrogation, and Confessions
July 13 DNA Saves