1. Introduction

This study involved developing sex offender treatment manuals and a Korean version of Multidimensional Inventory of Development, Sex and Aggression(MIDSA; Knight, 2009). It also attempted to suggest how current legal and systemic arrangements be changed for the developed manuals and assessment tools to be implemented appropriately.
In order to establish main principles and directions through which the Korean sex offender treatment programs can be improved and advanced, we reviewed the three main theoretical models for treating sex offenders: Relapse Prevention Model; Risk, Needs, Responsivity Model; Good Lives Model. Then we reviewed treatment components of sex offender group therapy, which was decreasing cognitive distortions and deviant sexual arousal that sex offenders have and increasing their victim empathy and socio-affective functioning and enforcing relapse prevention. Regarding the factors that affect therapy process we addressed characteristics of therapist and client, and therapeutic climate and approach. Lastly, based on the review made here, we derived several major principles that Korean sex offender treatment programs should pay attention to: considering Risk and Responsivity in administering sex offender treatment program; enforcing the motivation of sex offenders to participate in therapy and emphasizing the strength of sex offenders rather than their deficits.

2. Method

First, regarding the development and validation of a Korean version of MIDSA(K-MIDSA), we reviewed the full items of original MIDSA and decided to develop and validate the basic section(33 attitude scales and 4 lie scales) of the entire inventory considering the project period and budget. We translated the original items of the basic section of the MIDSA into Korean and reviewed the translated Korean items especially considering whether they are correctly translated with the sematic validity of the original items maintained. After several revisions of the translated items, we created two versions of K-MIDSA questionnaire, one for inmates and one for community people, and administered them to the two samples: 288 sex offenders and 190 community people.
Second, regarding the development of sex offender treatment manuals, we collaborated with Dr. William Marshall's team in Canada. Dr. Marshall is a widely accepted top expert in sex offender treatment and is well known for his lifelong dedication to research and treatment of sex offenders. We decided to develop 3 treatment manuals: generic manual, pre-treatment manual, full treatment manual. Generic manual included the theoretical models for treating sex offenders and the basic operating principles and procedures of sex offender treatment program. Pre-treatment manual adopted the motivational approach and contained the 8 topics that were presumed to be suitable for increasing the motivation for the participation in sex offender therapy. Full treatment manual involved the 21 topics that should be covered in the therapy sessions, and it divided the entire program into 3 phases: introduction; chronic criminogenic factors; self-management.

3. Results

1) K-MIDSA

In order to validate the K-MIDSA, we examined the reliability of its 33 subscales and its criterion validity. First, regarding the reliability, we examined the internal consistency of the 33 subscales with the sex offender group. Sexualization-related scales, paraphilia-related scales, pervasive anger-related scales, child molestation-related scales produced the good internal consistencies; Conbach's alpha ranged from .70 to .90. But, masculine adequacy-related scales, psychopathy-related scales, hypermasculinity-related scales showed relatively low internal consistencies.
We examined the criterion validity of the scales comparing the mean differences of the 33 subscales between the sex offender group and community group. Paraphilia-related scales, child molestation-related scales produced the statistically significant differences between them. Sex offenders showed higher scores in those scales. But, for other scales, we could not find the statistically significant differences. Then we compared the scores of the 33 subscales bewteen recidivists and non-recidivists. Recidivists showed higher scores in sexual compulsivity, paraphilia-related scales(except scatologia), child molestation-scales. We divided the sex offenders into 3 groups: rapists, child molesters, mixed type. Among the 3 groups, mixed typed sex offenders showed the highest scores in sexual compulsivity and paraphilia-related scales.

2) Sex Offender Treatment Manuals

A. Generic Manuals: major issues

All treatment programs described in the accompanying manuals follow the Principles of Effective Offender Treatment derived by Andrews and his colleagues from a series of meta-analyses (see Andrews & Bonta, 2006, for a summary). These meta-analyses allowed Andrews to identify the effective elements of offender treatment which they summarized as involving the appropriate application of three principles: Risk, Need and Responsivity. These three principles were shown to account for the effectiveness of programs for various types of offenders.
Most sex offender therapy uses group treatment. Group treatment have a number of unique features that are absent in individual work. Within groups, vicarious learning is clearly evident, as is role flexibility (being able to be both a help-seeker and a help-provider). Group treatment also results in the clear realization by clients that their problems are not unique, and it elicits empathetic behavior, and significant interpersonal learning.
If treatment providers decide to employ group therapy they must decide whether to run closed or open (often called rolling) groups. A closed group is one where all participants start and finish the program at the same time. This involves going through, in a fixed sequence, a series of components or modules that address each of the specific targets of treatment. An open-ended approach involves continuous intake such that as one participant completes all the targets and is discharged, a new client takes his place. As a consequence, each of the 8 or 10 clients will be at a different point of progress through the program. This means that boredom with any one target or exercise will be reduced, because each target will be addressed intermittently. An open-ended approach also provides the opportunity for the more senior group members to assist the newer clients. This approach also allows each client repeated opportunities to challenge and support newer members on the same issue. These repeated opportunities, when the more senior clients have continued to assimilate more and more of the program, allow for more sophisticated challenges, continuous vicarious learning, and repeated consolidation of the relevant learning.
The following schedules for treatment appear to be optimal for the appropriate operation of sex offender therapy. For institutionally-based treatment, depending upon how many sexual offenders are available, it may be possible to operate separate groups for high, moderate, and low risk offenders.
If this is the case, then the groups for the high risk offenders should run for approximately 10 months at the rate of three sessions per week with each session being 2.5 hours in duration; this results in 310 to 320 hours of treatment. For moderate risk offenders, approximately 6 months of treatment at the rate of three sessions of 2.5 hours per week is sufficient; this results in 188 to 190 hours of treatment. For low risk offenders, four months of treatment at the rate of two sessions per week of 2.5 hours duration should be enough; this results in 85 hours of treatment. When there is a limited number of offenders available for treatment at any time, then offenders at all levels of risk may be combined into one group. Under these circumstances a choice must be made about what to do with the low risk offenders. Having them in treatment for the same amount of time as the high risk offenders may result in over treatment which has been shown, particularly with low risk clients, to increase, rather than decrease offending (Lovins, Lowenkamp, & Latessa, 2009). There is no easy or obvious solution to this dilemma.
For treatment provided in an outpatient, community-based setting, clients are typically not so readily available for intensive treatment. Their jobs and other responsibilities limit the number of weekly sessions and efforts need to be made to provide treatment in the early evening. Again there are no empirical guidelines but most such programs run one or two 2.5 hour sessions per week. Our experience in operating a community outpatient program for 25 years, suggested that one session per week was not enough to maintain momentum and interest, so we recommend two sessions per week. In community settings, it is usually necessary to have offenders at all levels of risk in the same groups but where there are sufficient numbers available, separate groups for each level of risk is preferred.

B. Pre-treatment manual

It is evident from the literature, and clinical experience that when first offered treatment a significant number of sexual offenders either refuse to enter the program or have to be persuaded to enter and then do so with some reluctance. In addition, some who enter treatment later withdraw or are removed due to uncooperative behavior. Finally, among those who do complete treatment, some fail to satisfactorily attain the goals of the program, again due to a failure to fully engage. Although the positively-based approach to treatment in all programs should encourage engagement, it is evident that by the time sexual offenders enter treatment, some remain resistant. These sexual offenders do not profit from treatment because they simply go through the motions but fail to change. Marques, Day, Nelson and West (2005) describe this latter group of sexual offenders as those who did not get it (i.e., they failed to show the necessary changes). Marques et al. demonstrated that these offenders had higher post-release re-offense rates than did those who effectively participated in treatment (i.e., who got it). It is, therefore, necessary to provide a program that prepares sexual offenders for effective engagement in treatment. The following table summarizes the basic topics and the required session numbers for each topic to be addressed appropriately in the pre-treatment program.

C. Full Treatment Manual

The program is conceptualized as having three phases. Phase 1 is directed at motivating and engaging the clients, establishing the appropriate group climate, and beginning the development of an effective therapeutic relationship. Therefore the issues addressed in Phase 1 are meant to be non-threatening and it should be relatively easy for each client to succeed in all aspects of this phase. Phase 2 addresses the known criminogenic features (i.e., those that predict future offending and are potentially changeable). Finally, Phase 3 attempts to integrate what has been learned into a set of future plans.
The initial issues to be addressed including a deion of the rules of the group, an outline of the program, and an indication of the general findings of the effectiveness of sexual offender treatment, are covered in the first session. The subsequent targets to be addressed in the program include: the Good Lives Model, a cost-benefit analysis, an autobiography, the immediate offense triggers, self-esteem, coping and mood management, empathy, relationships, healthy sexuality, deviant sexual interests, and future self-development and self-management. The future self-development and self-management issues will include revisiting the Good Lives Model, identifying simple relapse prevention plans, forming support groups, and planning accommodation and employment goals. At the beginning of each session ask the clients how they have been since the last session. While this check-in is a nice way to begin each session, do not let this go on for more than five minutes except in unusual circumstances. Remember there are important issues to be addressed in each session.
When therapists are new to this work, the recommendation is to have no more than six offenders in the group. More experienced therapists can deal with 8 to10 offenders per group. Similarly while experienced therapists can run these groups on their own, novices should have a co-therapist. So long as therapists are competent and experienced, the gender of therapists appears to be irrelevant. However, there are obvious advantages to having a female and a male co-therapist; for example, many of the clients can profit from seeing a competent and empathic female, and having a male and female therapist offers the chance to model respectful gender relationships.
In most cases, clients are expected to complete each of the standard exercises in the order in which they are presented here. However, it may be necessary to alter the order in which a specific individual completes the exercises, or to combine assignments, due to factors affecting the client's ability to deal with some issues (e.g., past trauma, client dysfunctions, or co-morbid disorders). It is important to have group members complete their assignments between sessions then share them with the group. All participants are expected to initially provide positive feedback to the group member who is presenting his exercise before asking questions about the details of the exercise. Following these questions and group discussions, some clients may be asked to revise their assignment and present it again before moving onto their next exercise.
All clients should participate in all topics and all discussions about these topics but some clients, due to their present skills or lack thereof, may need more work on various issues and possibly less on others. The following table summarizes the basic topics and the required session numbers for each topic to be addressed appropriately in the full treatment program.