The purpose of the project was to investigate recidivism amongst juvenile and adolescent sexual offenders. Apart from determining the different relapse ratios amongst this research group, systematic risk factors were also identified. The main goal was to apply a number of Anglo-American actuarial risk assessment instruments on a German sample in order to test their limits and predictive validity in the local context.

Area of Re­search

In re­cent years the de­bate sur­round­ing sexu­al of­fend­ers has de­veloped in­to a highly di­vis­ive polit­ic­al is­sue. As a res­ult, re­search on sexu­al de­lin­quency and re­lapse has strongly in­creased. This re­search has however centered al­most ex­clus­ively on adult sexu­al of­fend­ers. It is only of late that ju­ven­ile and ad­oles­cent sexu­al of­fend­ers have also fallen with­in the scope of this re­search: pre­sum­ably as it is now abund­antly clear that ju­ven­ile and ad­oles­cent sexu­al of­fend­ers carry out a sig­ni­fic­ant num­ber of sexu­al of­fences. More than any oth­er char­ac­ter­ist­ics, the twin bio­lo­gic­al factors of age and sex af­fect these crim­in­al of­fences. Moreover, ju­ven­ile and ad­oles­cent sexu­al of­fend­ers rep­res­ent a ser­i­ous prob­lem for so­ci­ety as there is a real danger that their ac­tions will de­vel­op in­to a pat­tern of life-long de­vi­ant be­ha­vi­or. That said, in deal­ing with young sexu­al of­fend­ers cru­cial psy­cho­lo­gic­al changes can still be achieved through ap­pro­pri­ate and timely in­ter­ven­tion.

Pro­ject Goals

For the first time in Ger­many the pro­ject ap­plied a num­ber of Anglo-Amer­ic­an ac­tu­ar­ial in­stru­ments to pre­dict re­cidiv­ism on young Ger­man of­fend­ers. The pre­dict­ive valid­ity of the in­stru­ments was meas­ured and com­pared in an ef­fort to de­term­ine how suc­cess­ful they were in im­prov­ing risk and re­cidiv­ism pro­gnoses in a Ger­man con­text.


The ret­ro­spect­ive design of the pro­ject ex­amined a sample group of ju­ven­ile and ad­oles­cent sexu­al of­fend­ers who were tried and con­victed in ac­cord­ance with Ger­man crim­in­al law and who were re­leased from cus­tody between 2000 and 2002. The in­vest­ig­a­tion was con­duc­ted through the use of data entry ques­tion­naire spe­cific­ally de­veloped for the re­search pro­ject. The ques­tion­naires re­lied on an ana­lys­is of the of­fend­er’s per­son­nel doc­u­ments, whereby bio­graph­ic­al and clin­ic­al data, the nature and ser­i­ous­ness of the of­fence (char­ac­ter­ist­ics of the of­fend­er and the vic­tim) as well as so­cial sur­round­ings were also con­sidered. A num­ber of dif­fer­ent ac­tu­ar­ial risk as­sess­ment in­stru­ments were used: the Ju­ven­ile Sex Of­fend­er As­sess­ment Pro­tocol II (J-SOAP-II; Pren­tky & Righthand, 2003) and the Es­tim­ate of Risk of Ad­oles­cent Sexu­al Of­fense Re­cidiv­ism (ERA­S­OR; Worling & Cur­wen, 2001). The ap­plic­ab­il­ity of Stat­ic-99 on young of­fend­ers was also tested. In ad­di­tion, the fol­low­ing in­stru­ments were ap­plied to meas­ure pre­dict­ive valid­ity in re­la­tion to gen­er­al and vi­ol­ent re­cidiv­ism: Youth Level of Ser­vice/Case Man­age­ment In­vent­ory (YLS/CMI; Hoge & An­drews, 2006), Manu­al for the Struc­tured As­sess­ment of Vi­ol­ence Risk in Youth (SAVRY; Bor­um, Bar­tel & Forth, 2003) and Hare Psy­cho­pathy Check­list: Youth Ver­sion (PCL: YV; Forth, Kos­son & Hare, 2003). A com­par­at­ive group was formed from vi­ol­ent ju­ven­ile of­fend­ers who have not com­mit­ted sexu­al as­saults. The com­par­is­on between sexu­al and vi­ol­ent of­fend­ers made it pos­sible to de­term­ine the spe­cif­ic char­ac­ter­ist­ics of ju­ven­ile and ad­oles­cent sexu­al de­lin­quency.

The study was con­duc­ted in a se­lec­ted num­ber of ju­ven­ile cor­rec­tion­al fa­cil­it­ies in Ger­many. The total sample con­sisted of 294 young of­fend­ers. Table 1 provides a break­down of the sample.

Table 1: Dis­tri­bu­tion of the sample

Ju­ven­ile Cor­rec­tion­al Fa­cil­itySexu­al Of­fend­ersVi­ol­ent Of­fend­ers
Ad­elsheim (Baden-Wurttem­berg)4141
Halle (Sax­ony-An­halt)66
Neuburg-Her­ren­wörth (Bav­aria)3333
Neustrel­itz (Mecklen­burg-Vor­pom­mern)2424
Raßn­itz (Sax­ony-An­halt)11
Rock­en­berg (Hesse)99
Sprem­berg (Branden­burg)44
Wies­baden (Hesse)1313
Wriezen (Branden­burg)55
The rates of re­cidiv­ism of the sample group were ob­tained with the help of of­fi­cial gov­ern­ment re­cords from 2008 so that the valid­ity of the dif­fer­ent in­stru­ments could be tested over an 80 month peri­od. Five dif­fer­ent re­cidiv­ism cri­ter­ia were covered: gen­er­al re­cidiv­ism, re­cidiv­ism with de­ten­tion, re­cidiv­ism with de­ten­tion of more than two years, vi­ol­ent re­cidiv­ism and sexu­al re­cidiv­ism. Table 2 provides a break­down of the rates of re­cidiv­ism.

Table 2: Rates of re­cidiv­ism

Re­cidiv­ism Cri­ter­iaSexu­al Of­fend­ersVi­ol­ent Of­fend­ers
Gen­er­al re­cidiv­ism79 %78 %
Re­cidiv­ism with de­ten­tion48 %39 %
Re­cidiv­ism with de­ten­tion of more than 2 years31 %23 %
Vi­ol­ent re­cidiv­ism49 %56 %
Sexu­al re­cidiv­ism11 %2 %


The Anglo-Amer­ic­an pro­gnost­ic in­stru­ments achieved a fair to good res­ult when ap­plied to the Ger­man sample. Of­fend­ers whom the in­stru­ments con­sidered high risk where in fact more likely to of­fend­er soon­er than those con­sidered low risk. The res­ults con­firm the pre­dict­ive valid­ity and ap­plic­ab­il­ity of the Anglo-Amer­ic­an pro­gnost­ic in­stru­ments in the Ger­man con­text.

Dis­ser­ta­tion Su­per­visor

The dis­ser­ta­tion was su­per­vised by Dr. Klaus-Peter Dahle [E-Mail] from the In­sti­tute of Forensic Psy­chi­atry, Char­ité – Uni­versitätsmed­iz­in, Ber­lin.