Publication Type: Session Paper
Abstract: Planning telemedicine projects with geographically distant partners opens the door to practical planning and human resource issues, both of which can serve to hinder a project if not appropriately accounted for. Researchers and practitioners need to prepare for differing utilization patterns, access issues, and responses from staff.

Such large-scale projects, however, enhance understanding of and provide insight to technology selection and strategic planning. In addition, multi-faceted telemedicine endeavors provide lessons on management, implementation, partnership development, effective patient care and staff training. Dealing with these planning challenges can seriously impact patient results, caregiver acceptance, costs, and other essential outcome variables.

When properly planned and administered, though, geographically distant telemedicine projects can offer extraordinary access to specialists, more affordable care, reduced patient and family member burdens, as well as faster healing and decreased hospitalizations.

Publication Type: Conference Paper/Unpublished Manuscript
Abstract: Linkage analysis is a crucial part of the investigative process when faced with a possible series of related offences. Key to linking offenses is establishing behavioral consistency (i.e. how consistently an offender displays the same or similar behaviors across their series). Recent empirical studies (e.g., Bateman & Salfati, in press; Salfati and Bateman, 2005) have found little evidence of behavioral consistency looking at either individual or groups of behaviors in serial homicide. Several factors have been proposed to account for offending behavior variability. One explanation stipulated within the clinical and investigative literature is that these changes are due to an escalation in violence across series (Ressler & Shachtman, 1992). While escalation regarding the increase in the amount and the degree of injury inflicted has been observed in other serial crimes, such as rape (Hazelwood et al., 1989), no empirical studies have examined escalation in serial homicide. Looking at patterns of change, such as escalation, rather than idiosyncratic behaviors may provide a more valid way of examining consistency across homicide series. The present study looks at escalation in the quantity and quality of violence exhibited across homicide series using Multidimensional Scaling. This study is a first step towards understanding the patterns of change in crime scene behaviors of serial murderers, which may have significant implications for linkage analysis.

Publication Type: Conference Paper/Unpublished Manuscript
Abstract: Many probation and parole departments utilize actuarial instruments to identify offender risk/need factors and to develop assessment-based supervision strategies. To evaluate the effectiveness of a specific risk/need assessment tool, it is important for departments to validate their instrument on their correctional population to determine if it is predictive of recidivism across the various offenders that they supervise. As part of the Travis Community Impact Supervision Initiative, the Travis County Community Supervision and Corrections department has focused on utilizing risk/need assessment appropriately as part of a larger “top to bottom” re-engineering of the department. In a previous evaluation of the Wisconsin Risk Instrument in Travis County Community Supervision and Corrections, the department found that this instrument predicted re-arrest and incarceration for both male and female offenders over a two-year follow-up period. The current study builds on this earlier study by examining the ability of the instrument to predict specific types of probation violations, re-arrest, and incarceration across various racial/ethnic samples. The discussion will focus on the study results, implications for future evaluations of risk/need assessment tools, and suggestions for developing appropriate data-driven supervision policies.

Publication Type: Conference Paper/Unpublished Manuscript
Abstract: Traditionally, research in crime and delinquency has focused on factors that predict criminal offending. Specifically, demographic and social-structural characteristics of offenders are well established in the extant literature. Several authors have noted, however, that similarities exist between victims and offenders. However, studies examining the role of social-structural characteristics of victims are scant. The current study examines two social-structural components, education and occupation, among the violent death population in Maryland. Using data from the Maryland Violent Death Reporting System [MVDRS], this study uses a sample of 3,235 violent deaths victims (968 homicides, 941 suicides, and 1326 deaths of undetermined manner) to examine the variation of education and occupation background across manner of death. Results and implications for policy are discussed.

Publication Type: Conference Paper/Unpublished Manuscript
Review Method: Peer Reviewed
Abstract: Coding of interviewer and respondent behavior within the fielded interview has become an important tool for evaluating survey questions. Cannell et al (1968) described a behavior coding system that has evolved into perhaps the one most commonly used for this purpose (Fowler and Cannell, 1996). Previous research (e.g., Oksenberg et al, 1991; Zukerberg et al, 1995; Edwards et al, 2002) has focused on issues such as the system's reliability, whether it works equally well with live monitoring and recorded interviews, the sample size needed for pretesting, and the number of codes needed to identify questions that present problems for interviewers and respondents. This paper will describe an extension of behavior coding to assess questionnaire functioning across language, as part of an evaluation of the degree to which translated versions exhibit cross-cultural equivalence. For this study, behavior coding was performed for several portions of the questionnaire administered to adults (via telephone) in the 2003 California Health Interview Survey. The sample of interviews to be coded was selected so as to include interviews conducted in English, Spanish, and Korean. Further, language of interview and respondent cultural group were independently varied, so that some Korean respondents were interviewed in English and some in Korean; Latino respondents were interviewed in both Spanish and English; all other respondents were interviewed in English. Bilingual staff applied a system of codes specifically adapted for use in cross-cultural applications, and coded interviewer and respondent behaviors from the recorded interview segments. The paper will describe the coding process, assess inter-coder reliability across languages, and compare the frequency of problem codes across cultural group and across languages of interview. In summary, the paper will discuss the usefulness of behavior coding as a potential means for assessing the cross-cultural functioning of survey questions.