The public perceives the criminal justice system, and sentencing in particular, as being too lenient (Brillon 1988; Broadhurst and Indermaur 1982; Doob and Roberts 1983, 1984, 1988; Hough and Moxon 1985; Roberts 1992; Sprott 1996). There are, however, few studies of public attitudes toward the conditional release (CR(2)) system specifically, and research on correctional personnel's attitudes toward this aspect of the legal system is virtually nonexistent. This is problematic insofar as there are several reasons why indices of CR attitudes for both the public and correctional personnel are important to obtain. First, public attitudes may be considered by policymakers and critical decision-makers such as judges when they render CR decisions (Gibson 1980; Gottfredson, Warner, and Taylor 1988). Second, public attitudes may be important in shaping reform of existing CR policies (Roberts 1988). Third, the implementation, operation, and success of community correctional procedures such as CR is dependent upon public support (Senese 1992; Sigler and Lamb 1995). Fourth, having an index of correctional personnel's (e.g., judges, case management officers [CMO's], National Parole Board [NPB] members) attitudes toward CR is important, as there is support that sentencing and release decisions are affected by correctional personnel's attitudes toward the acceptability of the number of offenders released on parole (Canadian Sentencing Commission 1987; Hogarth 1971; Solicitor General of Canada 1981). Overall attitudes toward CR have also been found to affect CMOs' recommendations to grant or deny parole (Samra-Grewal, Pfeifer, and Ogloff 2000). Despite the importance of having an index of CR attitudes both for the public, as well as correctional personnel, there is very little research in this area, and there is presently no existing standardized instrument to assess these attitudes. The purpose of this study was to create such an instrument, primarily for future use in research contexts (e.g., in studies examining public support for CR or the release decision-making processes of correctional personnel).
Factors influencing legal attitudes in general
A variety of demographic characteristics have been found to affect the public's legal attitudes. In the United States, white males from moderate to upper income households have showed stronger support for capital punishment (Keri and Vito 1991), as have republicans and westerners (Bohm 1991). Respondents who are white, male, married, older, or have higher incomes have been found to have more punitive attitudes toward sentencing (Blumstein and Cohen 1980). Conservatism has been found to be correlated with more punitive attitudes toward the police (Zamble and Annesley 1987), `three-strikes-and-you're out' laws (Applegate, Cullen, Turner, and Sundt 1996), and parole (Brillon, Louis-Guerin and Lamarche 1984).
In addition to demographic factors, it has been hypothesized that a lack of information about the criminal justice system may contribute to more negative attitudes toward offenders (Roberts 1992; Valliant, Furac, and Antonowicz 1994). In fact, the public's knowledge about crime and sentencing is primarily derived from media accounts and interpersonal source reports (e.g., from family and friends), which are often selective and nonrepresentative (Broadhurst and Indermaur 1982; Diamond and Stalans 1989; Doob 1985; Graber 1980; Roberts and Doob 1990, Skogan and Maxfield 1981; Sprott 1996; Tyler 1980). The public's overestimation of the rates and nature of crimes committed, as will be reviewed in the subsequent section, may also contribute to more punitive legal attitudes.
Existing research on CR-related attitudes
Although there is evidence of public support for the theoretical principles of CR, the limited research on this topic suggests that, overall, the public's attitudes toward existing CR practices are not favorable (Brillon et al. 1984; Cumberland and Zamble 1992; Environics Research Group Limited and Dorothy Aaron Research Limited 1989; Roberts 1988; Zamble and Kalm 1990). There are several limitations with the current CR attitude indices. First, there is no existing standardized scale to assess attitudes toward CR. Rather, CR attitude indices have largely consisted of only one or several questions. The questions that have been utilized in previous research on CR attitudes have also tended to be quite global; this is problematic given findings that specific questions tap more complex and less punitive legal attitudes on the part of the public than is revealed by general questions (Applegate, Cullen, Link, Richards, and Lanza-Kaduce 1996; Cullen, Skovron, Scott, and Burton 1990; Cumberland and Zamble 1992; Diamond and Stalans 1989; Doob and Roberts 1984; Flanagan, McGarrell, and Brown 1985; McCorkle 1993; Skovron, Scott and Cullen 1988; Stalans and Diamond 1990; Thomson and Ragona 1987; Zamble and Kalm 1990). Given the above, there is considerable ambiguity in the literature with respect to the types of offenders and forms of release toward which respondents' attitudes have been directed.
A second limitation with existing CR attitude research is that the public as a whole holds inaccurate knowledge of the rates, nature, and severity of crimes that are committed by offenders (Brillon et al. 1984; Diamond and Stalans 1989; Doob 1985; Doob and Roberts 1982; Graber 1980), as well as inaccurate perceptions of the release and recidivism rates of offenders who have been conditionally released (Roberts 1988). A third and related limitation in the research is that respondents in CR surveys have been found to define CR erroneously …