intentions directly affect one's future behaviors

Several general risk factors have been reliably correlated with violent behaviors among clients with psychosis. One of the most consistent symptoms associated with dangerous or aggressive behaviors is overall severity of psychosis (e.g., hallucinations, delusions; Swanson, 1994). Perhaps this risk is heightened in those clients with late onset of psychotic symptoms (Montanez, 2000). The perspective that active psychosis is linked to violent behaviors has also been reported in recent international studies from Germany (Erb, Hodgins, Freese, Mueller-Isberner, & Joeckel, 2001), Finland (Eronen et al., 1996a), Canada (Montanez, 2000), New Guinea (Pal, 1997), India (Rath & Dash, 1990), and Denmark (Gottlieb, Gabrielsen, & Kramp, 1987). Additional symptoms or impairments that have been correlated with harm toward others include severe neurological impairments, suspiciousness and hostility (Krakowski, Czobor, & Chou, 1999), general self-care and interpersonal problems (Flannery, Penk, Irvin, & Gallagher, 1998), antisocial traits, and male gender (Citrome & Volavka, 2000; Soyka, 2000). According to at least one study focused on violence toward others, the risk may be more pronounced in clients who are unskilled, uneducated, and unmarried (Glancy & Regehr, 1992). Perhaps because of an exacerbation of other psychiatric symptoms, a general de-inhibiting effect, or noncompliance with treatment tiffany key pendant (Soyka, 2000), the risk of violence seems to be greater among clients with psychosis who are also dealing with active substance abuse or dependence (Citrome & Volavka, 2001; Swanson, 1994).

However, most of the studies previously cited exhibit notable methodological limitations (Eronen, 1995). Sample sizes have usually been small, thus limiting statistical power during data tiffany jewelry on sale analyses (Eronen, Tiihonen, & Hakola, 1996), and researchers often used outdated (e.g., pre-DSM-IV [Diagnostic and Statistical Manual of Mental Disorders, 4th ed.; APA, 1994]) diagnostic criteria to identify valid samples. Second, the research we previously cited focused on studying violent behaviors (i.e., some form of physically aggressive or dangerous behavior toward others) rather than on homicidality per se. That is, most prior studies investigated correlates of criminal acts, assaults, and completed homicides that could have occurred for a variety of reasons. Because these studies did not undertake an investigation of homicidality-predicting clients' thoughts or intentions to harm an identifiable person for a specific reason-in many instances, the violent behaviors reported may not have been preceded by homicidality. One common limitation of prior research on homicidality and psychosis is that the majority of previous studies in the literature used a retrospective approach to predicting future violent behavior. That is, researchers have generally documented clients' past violent behaviors (e.g., by inspecting medical or legal files, or reviewing autopsy records) and attempted to use this information for prediction purposes. At best, this research design can lead to general information about clients' demographic characteristics or personal history. Although it may be helpful to correlate past violent behaviors with violence committed in the future, this method is not adequate when one's goal is training clinicians to understand and prevent underlying psychosocial symptoms that generate homicidal ideation and intent in clients.

For example, literature from social psychology highlights the fact that one's thoughts and intentions directly affect one's future behaviors. That is, an individual's behavior can be linked to an immediately preceding intention to act a certain way (Ajzen, 1985). Stronger intentions usually result in specific target behaviors more often than do weaker intentions (Ajzen, 1991). Research demonstrates that thoughts and intentions may account for 30% of the variance in one's future actions (Golhvitzer, 1999). However, only two studies found to date have evaluated risk factors related to homicidal thoughts and intent among elsa clients with psychosis. After reviewing the detailed interviews that licensed clinicians conducted with 223 inpatient clients who had schizophrenia, Schwartz, Petersen, and Skaggs (2001) found that more impaired global functioning and more severe manic and psychotic symptoms predicted immediate homicidal ideation and intent. A more recent replication study was conducted by Schwartz, Reynolds, Austin, and Petersen (2003). These authors used the same methodology as Schwartz et al. (2001) but evaluated additional risk factors for homicidality. Their results showed that current manic symptoms, substance abuse, more impaired global functioning, and male gender were each significantly correlated with more extreme homicidality among inpatient clients with schizophrenia.