Les phénomènes de vieillissement et de comportement d'échelle dynamique ont été observés dans le passé dans beaucoup de systèmes hors-équilibre, mais il n'y a pas de théorie générale pour la description de ce type de système. Un premier pas a été réalisé avec la théorie d'invariance d'échelle locale (LSI) qui tente d'indentifier une forme généralisée du comportement d'échelle dynamique spatio-temporel. Pour des systèmes avec un exposant critique z = 2, il était déjà connu comment traiter ce genre de problème. Dans cette thèse on propose une reformulation et un élargissement de la LSI aux systèmes avec z ? 2. On en déduit des règles de Bargmann généralisées et on discute les symètries dynamiques élargies d'une équation de Langevin avec z ? 2. Puis on établit un formalisme pour le calcul des fonctions de corrélations et de réponse hors-équilibre. Les résultats sont confirmés dans plusieurs modèles concrets. Deuxièmement on s'intéresse au comportement de vieillissement dans des systèmes de réaction-diffusion. On trouve des différences importantes par rapport aux systèmes magnétiques dans certaines relations des exposants. Nous montrons aussi pour deux modèles exactement solubles comment élargir la LSI pour z = 2 afin d'inclure des modèles non-linéaires. Finalement, nous considérons le comportement de vieillissement proche d'une surface dans un système magnétique. Les résultats montrent que les formes d'échelle qu'on trouve dans le volume restent valides, mais certains exposants et fonctions d'échelle sont différents des quantités de volume correspondantes. / Ageing phenomena and dynamical scaling behaviour have been observed in many out-of-equilibrium systems, but a general framework for the description of such systems is still missing. A first step in this direction is the theory of local scale-invariance (LSI), which attempts to identify generalised forms of spatio-temporal dynamical scaling. For systems with a dynamical exponent z = 2, it has already been known how to treat stochastic partial differential equations and the consequences have been verified in many explicit models. In this thesis a reformulation and extension of LSI for systems with z ? 2 is presented. We infer for the first time generalised Bargmann superselection and discuss extended dynamical symmetries of Langevin equations with z ? 2. We can establish a formalism for the calculation of non-equilibrium correlation -and response functions and the results are confirmed in several new model calculations. Secondly, the ageing behaviour in reaction-diffusion systems is investigated. Although the main features of ageing as seen in magnets are still valid, important differences in exponent relations are found. Explicitly, the contact process is studied through field-theoretical methods and two bosonic models are solved exactly. For the latter, we show how to extend LSI with z = 2 to nonlinear models. Thirdly, the ageing behaviour in semi-infinite magnetic systems close to the surface is considered. The results show that the general scaling picture known from infinite systems remains valid, but some ageing exponents and scaling functions differ from the bulk quantities.

Zugl.: Paderborn, Universiẗat, Diss., 2004.

Child pornography offenders (CPOs) are ever present in the criminal justice system, yet the research on this population of offenders is less advanced than in many other areas of corrections (Eke & Seto, 2012; Seto & Eke, 2005). In order to effectively manage CPOs, it is necessary to accurately assess their risk, and, where applicable, provide rehabilitation options targeted toward their criminogenic needs. The current study examined the both the Level of Service Inventory-Ontario Revision (LSI-OR) and a modified version of the Child Pornography Offender Risk Tool (CPORT-M) and their ability to predict child pornography (CP), sexual, violent, and general recidivism with a sample that included CPOs, other sexual offenders (SOs), and non-sexual offenders (NSOs), who are under the responsibility of the province of Ontario. The results from the ROC analyses that examined the LSI-OR with the recidivism variables, for the various groups of offenders, suggested that the LSI-OR has good predictive accuracy for general recidivism for all of the offenders, as well as good predictive accuracy of violent and sexual recidivism with only the SO and NSO groups. Further, it was found that the CPORT-M had good predictive accuracy for general recidivism among the CPOs. It is appropriate to use both the LSI-OR and the CPORT-M to assess risk of general recidivism with CPOs.

Universiẗat, Diss., 2003--Jena.

In the United States, the large number of incarcerated individuals presents heavy social and economic burdens. To lessen these strains, many criminal justice agencies utilize risk assessment to determine which individuals are at a higher risk of recidivating and allocate limited intervention resources accordingly. To ensure these interventions are being delivered to those persons most in need, these risk assessment instruments must be tested for predictive validity. The present research seeks to revalidate one such risk assessment tool, the Level of Service Inventory Revised (LSI-R), on an adult offender population of a Midwest state. Additionally, this research expands on previous LSI-R validation studies by assessing the predictive validity of the LSI-R on an understudied population, Native Americans. The analyses utilized in this research include univariate descriptive statistics, bivariate correlation, and Receiver Operator Characteristic/Area Under the Curve analysis. Results are presented. Policy implications and recommendations for future research are discussed.

This research examines the possible correlations between a computing student's learning preference and their academic success, as well as their overall satisfaction with their major. CS and IT seniors at BYU were surveyed about their learning preferences and satisfaction with their major. The research found that IT students who are more reflective in their learning preference tend to have higher grades in their major. Additionally, it found that student age and their parents' education level were significant players in their academic success. However, there were no correlations found between major satisfaction and academic performance.

Current literature suggests that the Level of Service Inventory (LSI) and its derivatives (LSI-R, LS/CMI, LSI-OR) are capable of predicting violent recidivism, even though they were not initially designed for this function (Girard & Wormith, 2004; Mills & Kroner, 2006). The purpose of this study was to generate violence prediction scales, based on items or subscales from the LSI-OR, using five different statistical techniques. These analyses were completed on the full construction sample, then the males and the females separately to determine how the scales differ from each other and what, if any, benefits would accrue from utilizing a gender-specific scale.<p> A cohort of 27,027 offenders who were released from custody or entered into community supervision over a one year period was included in the study. There was an average followup time of 4.4 years. In this sample there was a general recidivism rate of 36.0% and a violent recidivism rate of 11.3%. Fifteen violence prediction scales were generated that ranged in predictive validity from r = .139 to r = .214. The scale with the highest predictive validity was the 11 item scale created from the full sample using the item linear regression technique. The scale contained items indicating that history of assault, lack of education and anger management issues were related to violent recidivism. Risk levels were developed for this new scale to classify offenders from very low to very high risk.<p> Although there was little difference in the predictive validity of the generated scales, the stepwise multiple linear regression technique was identified as the most successful method of creating a tool for predicting violent recidivism. There was no increase in predictive validity when using the scale that was developed for just the females in the sample, although fewer items were consistently generated for females than males. Therefore the full sample item linear regression scale is recommended for the prediction of violent recidivism of both male and female offenders in the jurisdiction from which the data were collected. Future research directions may replicate this study in other populations and further analyze the gender differences in violent recidivism.

Risk assessment and case management are two important aspects of young offender corrections and reintegration. Evaluating the extent to which case management practices are guided by risk assessment is important because the impact of the risk assessment instrument cannot be adequately assessed if the instrument is not being applied as fully intended. Unfortunately, little research has been devoted to examining the use of risk/need instruments in offender case management. The purpose of the present study was to investigate the link between risk assessment and community case management of young offenders in Saskatchewan and whether adherence to the principles of risk, need, and responsivity vis-à-vis the Level of Service Inventory Saskatchewan Youth Edition (LSI-SK; Andrews, Bonta, & Wormith, 2001) is related to recidivism. <p>Risk assessment and case management data were collected for a total sample of 193 young offenders who were supervised by youth workers from the Saskatoon and Regina probation offices. The sample was followed up for an average of 644 days. The overall recidivism rate was 62.2% with no significant difference in recidivism according to office of supervision, sex, or ethnicity.<p>The LSI-SK total and seven of the subscale scores were significantly, positively correlated with recidivism. Results also indicate that the LSI-SK was being used to guide supervision intensity as well as interventions. Moreover, the present study found that adherence to the need principle was associated with reductions in recidivism. Appropriateness (defined as the presence of interventions for identified needs or absence of interventions for areas that were not identified as needs) correlated significantly with recidivism (r = - .214). Appropriateness was found to be a significant predictor of recidivism after controlling for ethnicity and length of follow up. For every appropriate intervention listed on the case plan, the likelihood of recidivism was reduced by 24%. In terms of inappropriate treatment, under treatment was significantly correlated with recidivism (r = .283) but over treatment was not. Under treatment was a significant predictor of recidivism after controlling for ethnicity and length of follow up. For every identified need that did not have a corresponding intervention, the risk of recidivism increased by 91%. Implications for case management and direction for future research are discussed.

Risk assessment and case management are two important aspects of young offender corrections and reintegration. Evaluating the extent to which case management practices are guided by risk assessment is important because the impact of the risk assessment instrument cannot be adequately assessed if the instrument is not being applied as fully intended. Unfortunately, little research has been devoted to examining the use of risk/need instruments in offender case management. The purpose of the present study was to investigate the link between risk assessment and community case management of young offenders in Saskatchewan and whether adherence to the principles of risk, need, and responsivity vis-à-vis the Level of Service Inventory Saskatchewan Youth Edition (LSI-SK; Andrews, Bonta, & Wormith, 2001) is related to recidivism. <p>Risk assessment and case management data were collected for a total sample of 193 young offenders who were supervised by youth workers from the Saskatoon and Regina probation offices. The sample was followed up for an average of 644 days. The overall recidivism rate was 62.2% with no significant difference in recidivism according to office of supervision, sex, or ethnicity.<p>The LSI-SK total and seven of the subscale scores were significantly, positively correlated with recidivism. Results also indicate that the LSI-SK was being used to guide supervision intensity as well as interventions. Moreover, the present study found that adherence to the need principle was associated with reductions in recidivism. Appropriateness (defined as the presence of interventions for identified needs or absence of interventions for areas that were not identified as needs) correlated significantly with recidivism (r = - .214). Appropriateness was found to be a significant predictor of recidivism after controlling for ethnicity and length of follow up. For every appropriate intervention listed on the case plan, the likelihood of recidivism was reduced by 24%. In terms of inappropriate treatment, under treatment was significantly correlated with recidivism (r = .283) but over treatment was not. Under treatment was a significant predictor of recidivism after controlling for ethnicity and length of follow up. For every identified need that did not have a corresponding intervention, the risk of recidivism increased by 91%. Implications for case management and direction for future research are discussed.

Current literature suggests that the Level of Service Inventory (LSI) and its derivatives (LSI-R, LS/CMI, LSI-OR) are capable of predicting violent recidivism, even though they were not initially designed for this function (Girard & Wormith, 2004; Mills & Kroner, 2006). The purpose of this study was to generate violence prediction scales, based on items or subscales from the LSI-OR, using five different statistical techniques. These analyses were completed on the full construction sample, then the males and the females separately to determine how the scales differ from each other and what, if any, benefits would accrue from utilizing a gender-specific scale.<p> A cohort of 27,027 offenders who were released from custody or entered into community supervision over a one year period was included in the study. There was an average followup time of 4.4 years. In this sample there was a general recidivism rate of 36.0% and a violent recidivism rate of 11.3%. Fifteen violence prediction scales were generated that ranged in predictive validity from r = .139 to r = .214. The scale with the highest predictive validity was the 11 item scale created from the full sample using the item linear regression technique. The scale contained items indicating that history of assault, lack of education and anger management issues were related to violent recidivism. Risk levels were developed for this new scale to classify offenders from very low to very high risk.<p> Although there was little difference in the predictive validity of the generated scales, the stepwise multiple linear regression technique was identified as the most successful method of creating a tool for predicting violent recidivism. There was no increase in predictive validity when using the scale that was developed for just the females in the sample, although fewer items were consistently generated for females than males. Therefore the full sample item linear regression scale is recommended for the prediction of violent recidivism of both male and female offenders in the jurisdiction from which the data were collected. Future research directions may replicate this study in other populations and further analyze the gender differences in violent recidivism.