Psychometric Properties of The Services Matching Instrument: An Investigation of Reliability, Validity, and Factor Structure
MetadataShow full item record
Individuals with mental illness are overrepresented in the criminal justice system and are more likely to end up incarcerated than to receive appropriate psychiatric treatment (Abramsky & Fellner, 2003). Historically, the treatment of these offenders has primarily focused on factors related to their mental illness; however, a series of studies has shown that these individuals share psychiatric risk factors with individuals who are not criminal justice involved (Morgan et al., 2010; Gross & Morgan, 2013) and criminal risk factors with individuals who are not mentally ill (Morgan et al., 2010; Gross & Morgan, 2013; Wolff et al., 2013; Wilson et al., 2014). This highlights the importance of a holistic approach to effective treatment that focuses on both criminogenic risk and mental illness, as these individuals also have aspects of criminality that are supportive of a criminal lifestyle (Draine et al., 2002; Moran & Hodgins, 2004; Morgan et al., 2010). In order to effectively treat criminal justice involved persons with mental illness, mental illness and criminal risk must be assessed and identified as potential treatment targets (Morgan et al., 2012). Although measures of criminal risk and mental illness exist independently, the Services Matching Instrument (SMI; Morgan, Kroner & Mills, 2015) is the first measure to integrate assessment of both of these constructs into one concise and holistic measure. The SMI was developed to target the dual constructs of mental illness and criminality with the intent of matching the needs of criminally justice involved persons with mental illness with appropriate services. Specifically, the SMI is designed to assess psychiatric symptomology, criminal history, antisocial attitudes and associates, social networking, social functioning, substance abuse, negative affect, and traumatic history. Additionally, the SMI contains an infrequency scale composed of rarely endorsed items to provide information about the individual’s test taking attitudes. Previous research developed the measure and provided preliminary evidence for reliability (i.e., internal consistency, test re-test reliability) and validity. In a three study series, the SMI was reconfigured and reduced from 197 to 94 items with statistical and theoretical support. The SMI showed to be a reliable measure over time with an overall test re-test reliability coefficient of .88, and acceptable internal consistency. Furthermore, in a preliminary study of validity five of the SMI scales demonstrated a medium to large strength of association with respective reference measures (Olafsson, 2017). The purpose of the current study was to build on previous research and provide further support for the reliability, validity, and factor structure of the measure. The current study re-examined the internal consistency, and examined the concurrent, convergent, and discriminant validity of the scales. Results indicated that the SMI subscales demonstrated acceptable levels of reliability and overall demonstrated evidence for concurrent, convergent, and discriminant validity. The current study supported a four factor model for the SMI, with mental illness, substance use, criminalness, and other (i.e.., compilation of risk/need factors) each representing a factor. Results of the current study support the use of the SMI for service provision and treatment planning for criminal justice involved persons with mental illness.