Results Description
Data Analysis
A point correlation bi-serial was utilized in running the examination of the existing relationship amid the independent and variables that are quasi-independent. This was the incorporation of the LSI-R general sub scores and entire scores in the inclusion of scores for depression, anxiety, features of borderline as well as not supporting scores scales. The categorical dependent variables that were utilized in the research were conducted to establish whether the person was arrested after two years. A chi- square was additionally run in this research as an examination of the existing relationship amid all the categorical variables (DeCoster, 2004). This was done in order to establish whether the utilized subject held domestic abuse history or any other history that relates to sexual assault. The dependent variables were a representation of whether the female offenders got arrested within two years after being released from prison into the community.
Descriptive quantitative study’s statistics was used in the study which was obtained from meta-analysis. Meta-analysis was useful in generating maximum data that was necessary for the justification of the research (DeCoster, 2004). The utilization of meta-analysis as a design study helped in retrieving data in regard to the female offenders from the existing national and federal databases and the recent cases that have been conducted in the verification of LSI-R reliability and accuracy in the prediction of female offenders risk. His analysis was conducted for all the offenders that were established through the databases in order to increase the efficiency of the study. In order to determine the particular profiles of all the offenders, the study utilized a multivariate analysis in order to determine the general variation (DeCoster, 2004). After the scores correlations were calculated the study utilized sequential regression analysis design as a tool of prediction on the accuracy and reliability of LSI-R assessment in the prediction of female offenders scores. The analysis utilized age as a major study control variable which helped in controlling the result and increasing the general efficiency of the study.
The LSI-R accuracy was determined to be higher with the use of a designed approach as companies to the utilization of a mixed gender approach. This is because a mixed approach influences the generation of increased data thus increasing the probability of losing reliability and accuracy. It was thus established that LSI-R is not always accurate in predicting female offenders risk mainly because in most cases it utilizes a mixed approach for both genders as well as increased sentences range thus decreasing its ability (DeCoster, 2004).
Results Description
Means and standard deviations on the LSI-R general and sub-scores for all the normative groups were provided. Community offenders in the elements scored more as compared to custodial and community in the association of custodial offenders (Lowenkamp, Bechtel, 2007). This was particularly higher in sub scores of LSI-R such as criminal history, accommodation, finances drug issues that were mainly based o community offenses conducted by the female offenders. This was an inclusion of the total LSI-R score for the female offenders. For all the female offenders the differences can thus be termed to be very apparent particularly in the sub-scores scales of drug issues, attitudes or orientation, criminal history, employment, and education as well as the whole LSI-R scores.
The variance of analyses was essential in demonstrating that LSI-R general score does not hold any sex differences in all the indications. However, even with these results, there were crucial sentences order distinctions that were established. The analysis of the study thus established that female community offenders scored significantly less as compared to custodial offenders and the association of both custodial community offenders. Multivariate variance analysis on the subscales of LSI-R helped in the indication of the primary implications differences in LSI-R risk prediction in regard to both genders (Lowenkamp, Bechtel, 2007). In addition, this revealed a difference in sentences orders in the context of history criminal, employment, and companions. The analysis thus made the indication that community female offenders scored a bit lower than those under the grounds of custodial in combination of community offenses.
Bivariate correlations were utilized in the examination of the association of reoffending to the general score of LSI-R and the subscales of sentences order as well as gender. The apparent result was that there was a significant different in the context of re-offenses and age correlations for the female offenders particularly as well as male offenders across the scores and as well as the order of sentences. This was visible in total LSI-R scores and correlation of repeating offenses which held the highest coefficients of correlations that occurred for female offenders custodial being represented by r= 20 and that of male offenders with the inclusion of female community offenders being depicted by Rs =18. V repeating offenses and subscales correlations established that the criminal history sub-score was responsible for the production of the highest correlations with the repeat of offenses being apparent across the whole gender distinction and orders offenders sentencing. In the close follow-up was the subscale of education as well as employment. This was then followed by companions, accommodation and the issues of the drug for the female offenders in general. For the female, the occurrence of the LSI-R sub-scores did not end here as the highest coefficient were yielded by the scale of orientation and attitude with leisure and finance closing the scales which was a representation of the community in the combination of custodial and custodial offenses in their personal representation.
From the result, it is clear that LSI-R is not accurate in the general prediction of female offender’s risk. LSI-R scores rise is also linked to increased female likelihood of repeating particular offenses. With the replacement of LSI-R general score with LSI-R sub-scores, the subscale of LSI-R that is criminal history was indicated to be a very constant predictor of the occurrence of repeating offenses in the context of sentences order and gender with a predictive high score of an increased likelihood of reoffending. The LSI-R sub score of education and employment was established to be a major predictive in offenses repeat for all the involved male offenders with leisure and accommodation being the major predictive for custodial as well as community female offenders in general. Based on the study custodial and community offenders established the company and drugs issue scales in the respective nature as the reoffending predictive (Lowenkamp, Bechtel, 2007). Employment subscale was thus established to be an apparent risk predictor of females risks and the occurrence of e offending by the female offenders. Female’s offenders under the custodial area apparent predictive were education and in other cases, accommodation, as well as companions scales, were utilized as community offenders predictive.
References
DeCoster, J. (2004). Meta-Analysis Notes. Retrieved August 5, 2009 from http://www.stat- help.com/notes.html
Lowenkamp, C.T., Bechtel, K. (2007). The predictive Validity of the LSI-R on a Sample of Offenders Drawn from the Records of the Iowa Department of Corrections Data Management System. Federal Probation. 71.3.
