It is important to note that the research to date consists of construction validation studies, and that the assessment scales and instruments are being further refined through additional research as the instruments are implemented in larger settings. Construction validation research took place between 1999 and 2007 in four jurisdictions: 1) Colorado (prisons: N=154; parole: N=134); 2) Missouri (prisons: N=272; parole: N=150; probation: N= 310); 3) Minnesota (prisons: N=198; probation: N=233); and 4) Maui (probation: N=158). 

The six tables presented below provide the research results to date. Predictions to relevant outcome measures are shown in Table 1, below. Most importantly, row 3, shows the correlations between all gender-neutral and gender-responsive scales with the outcome, offense-related, variable examined in each study. Results are well within findings for the more well-established gender-neutral risk/needs assessments.

In Table 2, below, the predictive validity of the individual gender-neutral and gender-responsive scales are presented for all samples.

In Table 3, below, the construct validity findings of the gender-responsive scales is compared across the sites.

In Table 4, below, the preliminary inter-rater reliability findings are presented.

Table 5, below, shows the internal consistency findings for the interview scales.

In Table 6, below, the internal consistency findings for the survey scales are presented.

Further details and psychometric measures are shown in a summary report prepared for the National Institute of Corrections (NIC) and appear under the publications link on this page or on the NIC web page (www.NICIC.org). Construction validation research is not considered to be the final step in test construction. As such, the instruments are undergoing further validation research and refinements as they are implemented for use in large correctional agencies.