Psychiatry, Psychology and Law Discretionary ...

This article was downloaded by: [University of Nevada - Reno] On: 27 March 2014, At: 14:32 Publisher: Routledge Informa Ltd Registered in England and Wales Registered Number: 1072954 Registered office: Mortimer House, 37-41 Mortimer Street, London W1T 3JH, UK

Psychiatry, Psychology and Law Publication details, including instructions for authors and subscription information: http://www.tandfonline.com/loi/tppl20

Discretionary Release Decisions of Actual and Mock Parole Board Members: Implications for Community Sentiment and Parole Decision-Making Research a

Samuel C. Lindsey & Monica K. Miller

a

a

University of Nevada , Reno Published online: 09 Nov 2011.

To cite this article: Samuel C. Lindsey & Monica K. Miller (2011) Discretionary Release Decisions of Actual and Mock Parole Board Members: Implications for Community Sentiment and Parole Decision-Making Research, Psychiatry, Psychology and Law, 18:4, 498-516, DOI: 10.1080/13218719.2011.625619 To link to this article: http://dx.doi.org/10.1080/13218719.2011.625619

PLEASE SCROLL DOWN FOR ARTICLE Taylor & Francis makes every effort to ensure the accuracy of all the information (the “Content”) contained in the publications on our platform. However, Taylor & Francis, our agents, and our licensors make no representations or warranties whatsoever as to the accuracy, completeness, or suitability for any purpose of the Content. Any opinions and views expressed in this publication are the opinions and views of the authors, and are not the views of or endorsed by Taylor & Francis. The accuracy of the Content should not be relied upon and should be independently verified with primary sources of information. Taylor and Francis shall not be liable for any losses, actions, claims, proceedings, demands, costs, expenses, damages, and other liabilities whatsoever or howsoever caused arising directly or indirectly in connection with, in relation to or arising out of the use of the Content. This article may be used for research, teaching, and private study purposes. Any substantial or systematic reproduction, redistribution, reselling, loan, sub-licensing, systematic supply, or distribution in any form to anyone is expressly forbidden. Terms &

Downloaded by [University of Nevada - Reno] at 14:32 27 March 2014

Conditions of access and use can be found at http://www.tandfonline.com/page/termsand-conditions

Psychiatry, Psychology and Law Vol. 18, No. 4, November 2011, 498–516

Discretionary Release Decisions of Actual and Mock Parole Board Members: Implications for Community Sentiment and Parole Decision-Making Research Samuel C. Lindsey and Monica K. Miller


University of Nevada, Reno

Parole board members (PBMs) decide whether or not to release prisoners on parole. Parole release decisions can have a significant impact on the public so community sentiment is important. This research investigates whether actual PBMs make different release decisions than mock PBMs (college students). Participants decided whether to grant parole to a mock prisoner. Participants listed the factors that influenced their decision and their perceptions of the offender. As compared to mock PBMs, actual PBMs were more likely to deny parole, especially when perceptions of the inmate were controlled. Even so, actual PBMs perceived the prisoner as more likable and credible than mock PBMs. Findings may result from the different experiences, knowledge, and attributions that actual and mock PBMs make. Findings suggest mock PBMs may not be adequate proxies for studying PBM decision-making. When adopting parole guidelines, correction authorities should also take into account factors mock PBMs considered important. Key words: decision making; expert versus nonexpert; parole board; prisoner; public opinion perceptions.

Parole board members (PBMs) make important decisions about the timing and conditions of a prisoner’s release; these decisions affect the prisoner’s and the public’s welfare (Burke, 2003). When offenders are released early on parole they serve the remainder of their sentences in the community supervised by correctional authorities (Petersilia, 2003). Early release was originally designed as a procedure to respond to prison overcrowding and as an incentive to encourage pro-social behavior among prisoners (Petersilia, 2003). Despite these good intentions, parole remains a controversial process because it potentially puts the community in danger if the prisoner recidivates. Many people have

come to believe that PBMs are too lenient and that prisoners are granted parole too early or often (Burke, 2003; Roberts & Stalans, 1997). Research has yet to fully investigate whether PBMs make decisions that are contrary to the decisions that members of the community would make, if they had the chance. Results will have important implications for community sentiment and community support for the parole system. This study also has implications for future research on legal decision-making. Both PBMs and juries make important decisions regarding prisoners that affect the public. Jury research is typically conducted on mock jurors (e.g., using college students

Correspondence: Samuel C. Lindsey, Interdisciplinary Ph.D. Program in Social Psychology, University of Nevada, Reno, Mailstop 300, Reno, NV 89557. Email: [email protected] ISSN 1321-8719 print/ISSN 1934-1687 online Ó 2011 The Australian and New Zealand Association of Psychiatry, Psychology and Law http://dx.doi.org/10.1080/13218719.2011.625619 http://www.tandfonline.com


Actual and Mock Parole Board Members rather than actual jurors). Although there has been much past research concerning whether student mock jurors make similar decisions to those of actual jurors (e.g., Bornstein, 1999), very little research has been conducted comparing college students and actual PBMs (but see Carroll & Payne, 1977). This study investigates whether future research on parole decision-making should use college students as proxies for studying actual PBMs. This is an important step in researching parole decisions. The current study asked both actual PBMs and mock PBMs (college students) to read a prisoner’s case summary and determine whether the prisoner should be released on parole. Participants listed the factors that influenced their decision and indicated their perceptions of the prisoner. Results determined the similarity of decisions arrived at and factors taken into account by the two groups. Parole Procedures and Training Parole release procedures differ from state to state in the United States and may include a form of mandatory, discretionary, or a combination of mandatory and discretionary release procedures (Petersilia, 2003). Mandatory release procedures (also known as determinate sentencing) involve prisoners who have completed all or a percentage of their sentence and are released back into the community without supervision (Petersilia, 2003). This process determines exactly when a prisoner will be released; the prisoner does not qualify to be released any earlier. In contrast, discretionary release procedures (also known as indeterminate sentencing) occur when states permit PBMs to use discretion when deciding whether to grant parole (Petersilia, 2003). A prisoner is sentenced to a certain number of years in prison, but could be released early if the parole board decides the prisoner is a good candidate for release. The discretionary powers granted to PBMs

499

range from permitting PBMs full release powers to limited release powers (e.g., New York does not give PBMs discretion in releasing violent felons) (Petersilia, 2003). As of 2008, 34% of parole/releasing authorities required mandatory release, 21.3% permitted discretionary release, and 44.7% used a combination of mandatory and discretionary release (Association of Paroling Authorities International, 2008). States individually determine what training they require of their PBMs. For example, Montana requires new PBMs to (1) be trained by current staff regarding current state and federal laws and rules related to parole; (2) participate in training regarding offender pathology, treatment, and supervision; (3) receive training on the organization of the department of corrections; and (4) be instructed on issues relevant to American Indians in the state (Montana Board of Pardons, 2009). Vermont PBMs follow Montana’s first, second, and third requirements, and in addition, are encouraged to attend nationally recognized PBM training conferences (Vermont Parole Board, 2007). National organizations have developed conferences to systematize the training of PBMs (e.g., National Institute of Corrections, Association or Parole Authorities International) (Burke, 2003). Whether new PBMs are trained by the state’s current PBMs or by a nationally recognized organization, the material consists of, at least in part, recommended national guidelines developed by the U.S. Parole Commission. National guidelines provide PBMs an actuarial framework based on recidivism research that assists them as they evaluate each case on an individual basis (United States Parole Commission, 2003b). Factors that Influence Parole Decision-Making Researchers have examined the decisionmaking of PBMs to understand factors


500

S. C. Lindsey and M. K. Miller

that influence their decisions (Petersilia, 2003). Personal characteristics (e.g., preferences, beliefs) of PBMs may assist with their decisions because PBMs are taught that their personal characteristics should inform their decisions. For example, the PBM training handbook of the Association of Paroling Authorities International instructs PBMs to consult their ‘‘personal values’’ (Burke, 2003, p. viii). Personal characteristics may also influence PBM decisions because PBMs are, like all humans, susceptible to bias. PBMs may employ selection bias to prioritize which factors of the prisoner’s past are most important to consider. PBMs may also use personal experiences to help them decide whether an offender may still be a risk to the community. The role of PBM personal values has received much attention from critics over the last 40 years because of the perception that PBMs were inconsistent, biased, and too lenient in their release decisions (Petersilia, 2003). Consequently, critics called for changes to parole procedures. Prior to the 1970s, most states in the United States utilized discretionary release parole procedures (Burke, 2003). Since then, many states have restructured their procedures to limit or remove discretionary powers of PBMs in favor of mandatory procedures (Burke, 2003). As of 2002, rather than having a parole hearing, three out of every four prisoners were released automatically after completing the mandatory time (Petersilia, 2003). Today, PBMs use risk-prediction instruments (e.g., Salient Factor Score) to assist with their decisions (Burke, 2003; United States Parole Commission, 2003a). Actuarial instruments serve as guidelines and were developed to uniformly reduce disparity among, and create tighter control over, discretionary decisions (Petersilia, 2003). Research shows that PBMs most often consider factors such as institutional behavior, crime severity, criminal history,

incarceration length, mental illness, and victim input (Caplan, 2007). This study past research by showing which factors actual PBMs consider relevant to parole release decisions, and compare them to the factors that mock PBMs consider. Public Perceptions of Parole Mandatory release procedures are assumed to reduce the opportunity for prisoners to be released early (Petersilia, 2003). Recent research, however, based on Bureau of Justice Statistics data has shown that, in states that have discretionary release, prisoners actually serve longer sentences than states having mandatory release (Hughes, Wilson, & Back, 2001; Stivers Ireland & Prause, 2005). For instance, prisoners in discretionary release states served 2 months longer than those in mandatory release states (Hughes et al., 2001). This research contradicts popularly held views by the public that discretionary release procedures are too lenient (Petersilia, 2003). In sum, the public prefers mandatory release (because of the belief that discretionary release leads to prisoners being released too early), but research shows that discretionary release actually leads to more punitive outcomes (prisoners stay in prisons longer). Some researchers suggest the discrepancy between public perception and actual practice of discretionary release is because of misleading information provided to the public. For example, Tonry (2000) suggests the public often misunderstand parole due to sensationalized information about the parole system provided by the mass media or political figures. Such information often contributes to the public forming negative opinions about parole based on stereotypes (Tonry, 2000). Specifically, the availability and representativeness heuristics may result in cognitive biases. The availability heuristic occurs when an individual is confronted with


Actual and Mock Parole Board Members a difficult cognitive task of judging probability or frequency and makes a prediction based on how easily certain information is recalled (Tversky & Kahneman, 1973). The representativeness heuristic occurs when an individual predicts an outcome intuitively perceived as representative of the evidence, but does not consider the reliability of the evidence or its actual probability (Kahneman & Tversky, 1973). For example, a viewer watches the nightly news describing a crime committed by a recently paroled offender. Later, that viewer is interviewed about her perceptions of parole procedures in general. Because the news story is quickly available (availability heuristic) she may recall the incident and think it is reflective of the majority of prisoners released on parole (representativeness heuristic) and respond that she thinks parole procedures are too lenient. The reality may be that the news highlighted a rare event and the woman misjudged the frequency that parole results in recidivism. This example demonstrates one possible explanation for low public support for discretionary release. Low public support for parole has a lengthy history. As early as national-level records are available measuring public sentiment toward parole, the public has perceived parole as needing to be stricter (71.7%) and more focused on punishment (Gallup Poll, 1937). This belief is similar to a 1994 poll; 82% of Americans favored or strongly favored ‘‘making it more difficult for those accused of violent crimes like murder and rape to be paroled’’ (Maguire & Pastore, 1995, p. 172). Polls about the public’s general perceptions of the parole system, however, may not be reflective of the public’s perceptions of how PBMs should deal with individual prisoners. For example, Cumberland and Zamble (1992) asked members of the Canadian public their beliefs toward parole and found 82% to think: ‘‘In general, the parole system is too lenient,’’ ‘‘somewhat

501

lenient,’’ or ‘‘much too lenient’’ (pp. 446– 447). However, when they provided participants a scenario describing specific parole-eligible offenders, 86% of participants paroled a first-time burglar with a high level of program involvement whereas only 28% paroled a violent personal offender with a previous criminal history and no program involvement. Participants appeared to vary in their punitiveness when asked about specific offenders and, in some cases, were more lenient and flexible than other members of the public when asked about offenders in general. Thus, researchers and policy makers who have interpreted public polls as a demand for stricter release procedures may be oversimplifying community perceptions that are in fact more complex. The public may think that PBMs should be stricter in general (because when asked they are thinking of violent offenders); but may also think that PBMs should be lenient on a case-by-case basis with certain types of offenders. It follows that an important contribution to the literature is to study the judgments of members of the public juxtaposed with judgments of PBMs, when faced with the same information about a prisoner. This will determine whether PBMs indeed make more lenient decisions than the public would make, when faced with the same decision-making task. The current research will also challenge the policy implications drawn from past parole decision-making research (i.e., to abolish discretionary release). Community Sentiment For legal systems to be perceived as legitimate authorities, it is important that they reflect the sentiment of the community (Tyler, 2006). Community members develop high regard for institutions they perceive to be legitimate and are more likely to accept and follow the institution’s decisions (Tyler, 2006). This may explain


502


why 15 states and the U.S. federal parole system abolished parole in the 1970s, 1980s, and 1990s. Other paroling authorities may recognize that community sentiment can be problematic: the presence of community sentiment does not necessarily mean it should be heeded (Finkel, 1995). For example, if the community does not have all the facts, they may not be able to make educated decisions regarding whether discretionary release should be abandoned, or what type of procedure should replace it. It is also possible that past opinion polls may have not asked the right kind of questions that accurately inform policy makers of the community’s true sentiment. This may explain why 16 other states still grant their PBMs full discretionary release powers. The perceived legitimacy of the paroling authorities may be affected by whether the community perceives PBMs make release decisions using factors that the community considers relevant. Indeed, the legitimacy of the parole system was questioned beginning in the 1970s when research revealed disparity among PBM decisions. Understanding what factors community members consider important to making release decisions may potentially inform paroling authorities of additional factors they should recommend as guidelines in the future. By adopting factors the public considers important, paroling authorities may increase their legitimacy. Using Students as Proxies for PBMs The PBM decision-making process has been questioned and criticized by researchers and the public for decades (Burke, 2003). Consequently, PBMs may be hesitant to assist researchers. PBM decisionmaking is also difficult to study because such research requires time, resources, and a relationship of trust between researchers and PBMs (Carroll & Payne, 1977). As

such, a need exists to determine whether another population can serve as proxies for PBMs. Using members of the public acting as mock PBMs may be an option. The field of jury decision-making faces a similar dilemma. Because jury researchers often have difficulty studying actual jurors, they often study members of the public (often college students) acting as mock jurors. There is some criticism of whether mock jurors are accurate proxies for actual jurors, yet some research has shown that college students do not uniformly differ from actual jurors in most types of jury decisions (Bornstein, 1999). Thus, for some fields, studying proxies may be an inexpensive and convenient alternative. PBMs are often selected from individuals with relevant training in corrections and interest in working with the prisoner population (Burke, 2003). It is possible that college students interested in criminal justice may be likely to serve as future PBMs. If college students acting as mock PBMs make similar decisions to actual PBMs, students may be a more convenient proxy for actual PBMs. Understanding whether mock PBMs decisions can substitute for actual PBM decisions is reflective of a larger discussion of whether studying non-experts can inform research of how experts make decisions. Expert versus Non-Expert Decision-Making One of the difficulties with conducting research using mock PBMs is that they do not have the training of actual PBMs (described above). As such, actual and mock PBM decisions may differ. Other fields of research (e.g., law, agriculture) also study whether the decision-making of non-experts (e.g., college students) can inform our understanding of experts (e.g., judges). Some fields have determined that both groups make similar decisions although employing different processes; other fields have determined that both


Actual and Mock Parole Board Members groups make very different decisions (Shanteau & Stewart, 1992). Experts and non-experts often employ different strategies for making decisions (Shanteau & Stewart, 1992). Experts and non-experts often differ in confidence but not necessarily in accuracy on decision tasks (Shanteau & Stewart, 1992). For example, experienced agricultural judges who judged the quality of wheat grain to determine which grain would produce the highest crop yields were more confident but no more accurate than less-experienced agricultural judges (Trumbo, Adams, Milner, & Schipper, 1962). Similar results have been observed among corn judges and other types of agricultural judges (Shanteau & Gaeth, 1981). Researchers have also examined how experts such as courtroom judges make decisions and found that judges only considered one to three factors when setting bail (Ebbesen & Konecni, 1975), whereas jurors rely on additional and irrelevant information (Schwarz, Strack, Hilton, & Naderer, 1991). This is similar to other experts who better discriminate important from irrelevant information compared to non-experts (Shanteau, 1992). Even so, jurors and judges typically make similar decisions (Kalven & Zeisel, 1966). Courtroom judges have also been found to have more procedural knowledge than jurors (Kardes, Muthukrishhnan, & Pashkevich, 2005) and that knowledge helps judges better predict future patterns (Phillips, Klein, & Sieck, 2004). Thus, experts and non-experts may differ in the process, even if they do not differ in the actual decision. As far as we are aware, no study has examined differences between experts and non-experts in the domain of parole decision-making. Carroll and Payne (1977), although not specifically studying parole release decisions, offer the closest look at differences between PBMs as experts and college students as non-experts in their study examining judgments of

503

seriousness about crime and risk of recidivism. Carroll and Payne recruited expert parole decision makers (i.e., 24 PBMs, parole board directors, parole supervisors) and non-experts (i.e., 64 college students) and found that non-experts were more likely than experts to make personal attributional biases when sentencing a criminal to prison. Attributional dimensions (internal vs. external, stable vs. unstable) of causes of the crime were found to be predictive of non-expert, but not expert decisions. That is, college students were likely to assign negative evaluations (higher ratings of seriousness, prison term, imprisonment to punish, recidivism) when they perceived internal attributions of the offender to be the cause of the crime. Some of the experts were also found to make attributions, although ‘‘more weakly’’ than the non-experts (p. 600). Carroll and Payne sum up their research by suggesting that non-experts may make good proxies for experts in situations in which decisionmaking tasks call upon only general knowledge and simple reasoning. However, when training and experience are involved and complex reasoning skills required, as is the case with parole decisions, non-experts may not be good proxies. Even though Carroll and Payne (1977) offer helpful suggestions as to whether nonexperts should represent experts, their recommendations may be limited because they only studied participants’ perceptions of offenders rather than release decisions. The current study will add to the findings of Carroll and Payne and further clarify whether mock PBMs make valid proxies for actual PBMs. Overview of Study The purpose of this study is to contrast the decisions and decision-making processes of actual PBMs with mock PBMs. The research will answer the following questions:

504



(1) Do actual and mock PBMs differ in their decision whether to parole a prisoner? (2) Do actual and mock PBMs differ in how they make parole decisions (e.g., in the factors they use and the complexity of their decisions)? (3) Do actual and mock PBMs differ in perceptions of a prisoner? No hypotheses are made because there is so little previous research comparing actual and mock PBMs. Actual and mock PBMs completed a study in which they read a case summary of a prisoner who was being considered for parole. Participants then indicated whether the prisoner should be released, the factor(s) influencing that decision, and their perceptions of the prisoner.

these, 67% were male and ages ranged from 37 to 77 (M ¼ 55, Mdn ¼ 54). Ethnically, 75% were White, 11% were AfricanAmerican, 8% were Hispanic, 3% were Asian/Pacific Islander, and 3% were other ethnicities. Mock PBMs were undergraduate students in introductory criminal justice courses. Recruiting from these courses was preferred because it was thought that students interested in criminal justice may be more likely than other students to seek employment in the field of corrections and subsequently be chosen as PBMs in the future. Of the 190 mock PBMs, 48% were male, and ages ranged from 18 to 33 (M ¼ 21, Mdn ¼ 20). Ethnically, 75% were White, 10% were Hispanic, 5% were Asian/Pacific Islander, 5% were AfricanAmerican, and 5% were other ethnicities. Materials and Procedures

Method Participants Participants included 54 actual and 190 mock PBMs (college students). Actual PBMs were recruited from 42 (excluding states that use only mandatory release). Contact information for actual PBMs was obtained from state parole board websites. Two-hundred forty recruitment letters were sent by email, mail, and fax to individual PBMs. In cases in which the identity of actual PBMs was not available, letters were sent to PBM supervisors or other officials asking for the letter to be forwarded to the PBMs. Actual PBMs were given the option of completing a survey online or requesting a paper survey. Follow up packets (i.e., recruitment letters, paper surveys, stamped/addressed envelopes) were mailed to participants who had not yet responded after two weeks. Email reminders were sent approximately 10 days later. Fifty-four actual PBMs participated (response rate of 23%), and of those, 36 PBMs completed the entire survey. Of

Participants read a 650-word case summary describing a mock prisoner, David Mitchum, who was eligible for parole. The case summary was written by the researchers based on an actual case summary provided by an actual PBM. Two actual PBMs reviewed and verified that the case summary was realistic of information included in actual case summaries (although the summary was more abbreviated than an actual case summary). The case summary stated that the prisoner, David Mitchum, had been convicted of second-degree murder at the age of 22 and sentenced to 25 years incarceration. Mitchum served 10 years in prison and was currently eligible for parole. The case summary discussed Mitchum’s criminal history, drug history, employment history, education, prior parole/probation revocations, institutional activities, programs completed, institutional disciplinary history, psychological evaluation, and case1 worker notes. Following the case


Actual and Mock Parole Board Members summary, participants indicated whether they would grant or deny parole to the prisoner, how certain they were of their decision (on a 5-point scale), what factors influenced their decision, and their perceptions of the prisoner. One qualitative and three quantitative measures addressed the three research questions of this study. These included a measure of parole decision certainty, factors influencing parole decisions, parole decision complexity, and perceptions of the prisoner. Parole decision certainty was measured by multiplying the participant’s parole decision (grant ¼ þ1, deny ¼ 71) with how certain participants were of their decision (1 ¼ very uncertain to 5 ¼ very certain). This yielded scores from þ5 to 75. Decision certainty is a common variable in measuring jury decision-making (also a group decision). Participants were asked the open-ended question: What factors influenced your decision to grant or deny parole to the prisoner? Their responses were divided into ‘‘factors.’’ Participants often listed several factors that influenced a decision; thus, there were many more factors than participants. A factor could have influenced the participant’s decision to deny or grant parole. ‘‘Commonly mentioned factors’’ were identified and grouped together. The commonly mentioned factors were further grouped into ‘‘themes’’. The researchers discussed the commonly mentioned factors, practiced with a small subset of participant responses, made changes to the codebook (e.g., to better define the commonly mentioned factors), and then did the interrater reliability check. An interrater reliability indicated that the coders had a high level of agreement; Holsti’s coeffient was .98. A typology was created of themes and commonly mentioned factors that influenced parole release decisions. A commonly mentioned factor was required to be mentioned by at least

505

three different participants before it was included in the typology. Additionally, a commonly mentioned factor was included in the typology if it was listed by Caplan (2007) as one of the six common factors shown to affect parole board decisionmaking in past research. Parole decision complexity was measured by calculating the number of factors each participant listed that affected their parole release decision. Perceptions of the prisoner were measured on a 5-point scale. Participants were asked to rate (1 ¼ not at all, 5 ¼ very) the credibility, honesty, and likeability of the prisoner and rate (1 ¼ very negative, 5 ¼ very positive) their overall perceptions of the prisoner. Results Decisions of Actual and Mock PBMs Research Question One compared decisions of actual and mock PBMs. Of 190 mock PBMs, 72.1% denied and 27.9% granted parole. Of 36 actual PBMs, 86.1% denied and 13.9% granted parole. Chisquare analysis revealed this to be marginally significant (X2(1) ¼ 3.11, p ¼ .078). The second measure, decision certainty, was also marginally significant (t(217) ¼ 1.70, p ¼ .09), indicating that actual PBMs were more confident in denying parole (M ¼ 2.69) than mock PBMs (M ¼ 1.76). Because the two groups differed in perceptions of the prisoner (as reported below), an ANCOVA was conducted to control for these perceptions. The difference between actual and mock PBMs was significant (F(1,217) ¼ 9.60, p ¼ .002, Zp2 ¼ .042); actual PBMs (M ¼ 72.89, SD ¼ 2.82) were more likely to deny parole than mock PBMs (M ¼ 71.73, SD ¼ 3.24). Decision-Making Process of Actual and Mock PBMs Research Question Two concerned the decision-making processes of actual and


506


mock PBMs. A qualitative analysis established a typology of commonly mentioned factors that affected parole release decisions (see Table 1). Participants listed 822 factors that influenced their decisions to grant or deny parole. Researchers identified 24 commonly mentioned factors and 9 themes (see Table 2 for definitions and examples). When denying parole, actual PBMs listed 98 factors and mock PBMs listed 538 factors. When granting parole, actual PBMs listed 10 factors and mock PBMs listed 177 factors. Both actual and mock PBMs listed violent behavior in prison as being one of the most influential factors for denying parole (ranked second for actual PBMs and first for mock PBMs). Actual PBMs did not list 13 of the commonly mentioned factors that were listed by mock PBMs for granting parole. Nine actual PBMs (25%) and 12 mock PBMs (.06%) stated they did not have enough information to make a decision. These actual PBMs always denied parole, whereas, only about half of these mock PBMs denied parole. The top three factors listed by actual PBMs for denying parole were lack of attending relevant programs in prison, violent behavior in prison, and severity of the crime. The top three factors listed by mock PBMs for denying parole were violent behavior in prison, lack of demonstration of life improvement, and lack of behavior consistency in prison. Very few factors were listed by actual PBMs for granting parole; the top factor listed was that the prisoner had completed a sufficient portion of his sentence. The top three factors listed by mock PBMs for granting parole were a demonstration of life improvement, benefit of the doubt, and there was a tie between positively affecting others in prison and attendance of relevant programs in prison. These findings are presented in Table 1. Next, differences in complexity of decisions were analyzed. Actual PBMs listed

1 to 7 factors while mock PBMs listed 1 to 11 factors that affected their release decision. An independent-samples t-test revealed that actual PBMs (M ¼ 2.97, SD ¼ 1.73) listed significantly fewer factors than mock PBMs (M ¼ 3.76, SD ¼ 1.97; t(224) ¼ 2.24, p ¼ .026). Perceptions of the Prisoner Research Question Three concerned PBM type and perceptions of the prisoner. Perceptions of the prisoner (i.e., overall perception, honest, credible, likeable) were highly correlated (ps 5 .001). A one-way MANOVA with PBM type as IV and the four perceptions as DVs was significant (F(4, 218) ¼ 2.537, p ¼ 0.041, Zp2 ¼ .044). Univariate analysis indicated that this effect was driven by perceptions of the prisoner’s likeability and credibility. Actual PBMs perceived the prisoner as more likeable (M ¼ 3.09, SD ¼ .658) than mock PBMs (M ¼ 2.70, SD ¼ .780, p ¼ .006). Similarly, actual PBMs perceived the prisoner as more credible (M ¼ 2.66, SD ¼ .639) than mock PBMs (M ¼ 2.37, SD ¼ .737, p ¼ .030). Discussion One purpose of this research was to investigate whether community sentiment is reflected in the release decisions of PBMs. A second purpose of this research was to investigate whether college students (as non-experts) make adequate proxies for PBMs (as experts) on a decision-making task. This study sought to accomplish these purposes by answering three research questions. This section discusses the findings and limitations of the current research, and the implications of this study for future research. Findings Research Question One. Research Question One concerned whether actual PBMs and

Note: *Indicates a tie in ranking with another factor.

Miscellaneous

Procedural

Insufficient information

Affected others

Future behaviors

Past behaviors

Skills & knowledge in prison

Attitudes & beliefs in prison

Violent behavior Positive behavior (or neutral) Insufficient time of positive behavior Mental health problems (or lack) Consistency (or lack) Anger/aggression Life improvement (or lack) Relevant programming (or lack) Education (or lack) Criminal history (or lack) Crime severity (or lack) Current age Employment history (or lack) Parole revocation (or lack) Drug history (or lack) Risk to recidivate Parole-release plans (or lack) Interpersonal other Victim input (or lack) Insufficient case detail No reason Insufficient no. Parole hearings Sentence served (or lack) Benefit of doubt

Commonly mentioned factors 14 0 7 0 3 4 4 15 3 5 11 1 1 0 3 2 2 0 1 9 2 3 7 0

Freq. 2 11* 5* 11* 8* 7* 7* 1 8* 6 3 10* 10* 11* 8* 9* 9* 11* 10* 4 9* 8* 5* 11*

Rank

Actual PBMs

97 0 21 2 84 30 97 12 46 31 24 0 9 1 14 22 3 22 0 6 1 1 11 4

Freq. 1* 18* 8 16 2 5 1* 10 3 4 6 18* 12 17* 9 7* 15 7* 18* 13 17* 17* 11 14

Rank

Mock PBMs

Deny parole

0 1 0 0 0 0 0 1 0 0 0 1 0 1 1 0 0 0 0 0 2 0 3 0

Freq. 4* 3* 4* 4* 4* 4* 4* 3* 4* 4* 4* 3* 4* 3* 3* 4* 4* 4* 4* 4* 2 4* 1 4*

Rank

Actual PBMs

13 8 3 2 4 2 47 17 5 3 6 8 0 2 6 2 0 17 0 6 1 0 7 18

Freq.

4 5* 10* 11* 9 11* 1 3* 8 10* 7* 5* 12* 11* 7* 11* 13* 3* 13* 7* 12 13* 6 2

Rank

Mock PBMs

Grant parole

Typology of themes and commonly mentioned factors why parole was denied or granted by actual and mock PBMs.

Behaviors in prison

Themes

Table 1.


Actual and Mock Parole Board Members 507

Miscellaneous

Procedural

No reason Insufficient no. parole hearings Sentence served (or lack) Benefit of doubt

Victim input (or lack) Insufficient case detail

Affected others

Insufficient information

Parole-release plans (or lack) Interpersonal other

Future behaviors

Anger/aggression Life improvement (or lack) Relevant programming (or lack) Education (or lack) Criminal history (or lack) Crime severity (or lack) Current age

Employment history (or lack) Parole revocation (or lack) Drug history (or lack) Risk to recidivate

Attitudes & beliefs in prison Skills & knowledge in prison Past behaviors

Insufficient time of positive behavior Mental health problems (or lack) Consistency (or lack)

Positive behavior (or neutral)

Violent behavior

Commonly mentioned factors

No reason is cited. Number of hearings (e.g., denying parole because it is the first hearing). Insufficient or a sufficient sentence has been served. References about the prisoner deserving a second chance including positive thinking on behalf of the respondent regarding inconclusive or insufficient case summary information (e.g., ‘‘I don’t have details of the fight in 2006, but he might have been provoked’’).

Misconduct by the prisoner while in prison which contributes to a dangerous environment (e.g., adjustment to prison, fighting, violence). Positive behavior while in prison including helping others or cooperating with the authorities (e.g., helping in prison, relatively little trouble). Needing to observe more time of positive behavior (e.g., more time to show proof of sincerity to change). Potential for (or lack of) mental, psychiatric, or psychological health problems. Consistency or inconsistency of thought and behavior (e.g., ‘‘even though he goes to counseling he still got into a fight’’). Reference to anger other than fighting or violence. Motivation or positive or negative desire for life improvement or lack thereof. Vocational training, anger management, drug counseling, substance abuse counseling, etc. Obtaining education while in prison or failure to finish education. Reference to prior crimes. Reference to current offense. Reference to his current age or indirect references to current age by mentioning he was much younger when committed prior offenses. Reference to employment or jobs. Reference to past paroles being revoked or not. Past or the possibility of current problems with drug abuse. Reference to recidivating. The word ‘‘recidivate’’ may be used or some other reference to future criminal behavior (e.g., ‘‘he will just be back’’). Goals or plans for his life if and when he is released on parole. When the prisoner’s relationship to other people is directly mentioned, positively or negatively. Information by the victim regarding whether the inmate should be released. Unable to accurately or confidently provide release decision without more information.

Definition and examples

Definitions and examples of themes and commonly mentioned factors from typology of parole release decisions.

Behaviors in prison

Themes

Table 2.


508 S. C. Lindsey and M. K. Miller


Actual and Mock Parole Board Members mock PBMs differ in their release decisions. Prior research has shown that the public perceives discretionary release procedures to be too lenient (Burke, 2003; Roberts & Stalans, 1997). This suggests that members of the public may make stricter decisions compared to PBMs. The current research refutes this perception. It is a bit surprising that actual PBMs (experts) were stricter than students (nonexperts) since other research on legal decision-making shows that judges (experts) are less punitive than jurors (nonexperts) (Kalven & Zeisel, 1966). However, this could be partially due to the difference in decisions (e.g., a verdict versus parole release). The difference could also be due to differences in experience. Actual PBMs have experience and training and thus know the ‘‘risk factors’’ for recidivism. They may have a different (and stricter) threshold for determining that the prisoner should be released. An untrained, inexperienced mock PBM likely has little basis for a decision. They may assume that, if the prisoner is eligible and has no major infractions, that they should be released – resulting in a lower threshold than actual PBMs. Research Question Two. Research Question Two concerned the decisionmaking processes of mock and actual PBMs. In addition to making different release decisions, actual and mock PBMs generally arrived at their decisions very differently. There were, however, some factors both groups listed less or more influential to their decisions. Decision-Making Similarities. Actual and mock PBMs both listed violent behavior in prison as being a very influential factor for denying parole. Participants may view violence in prison as a likely predictor of the prisoner continuing violence once released. When denying parole, both actual and mock PBMs only occasionally mentioned: the prisoner’s current age, whether the prisoner had prior parole revocations,

509

demonstrated positive behavior in prison, had mental health problems, and whether victim input was provided. When granting parole, actual and mock PBMs infrequently mentioned: insufficient time of positive behavior, mental health problems, anger in prison, criminal history, employment history, risk to recidivate, postrelease plans, victim input, and number of parole hearings. Decision-Making Differences. Despite the similarities just mentioned, there was little agreement between actual and mock PBMs in the majority of factors in which parole was denied. Of particular note is the factor, lack of attendance of relevant programming, which actual PBMs listed most often when denying parole and mock PBMs mentioned much less often. Actual PBMs apparently consider participation in programs a key to preparing prisoners for integration back into the community. PBMs likely believe that institutional programming is able to rehabilitate prisoners; otherwise they would have difficulty justifying the release of prisoners. Actual and mock PBMs also differed in the majority of factors when granting parole. The most common factor actual PBMs listed for granting parole was that sufficient time had been served for the crime. Mock PBMs listed the prisoner’s desire to improve his life as the most commonly mentioned factor for granting parole. Actual PBMs did not list the majority of factors listed by mock PBMs when granting parole. Apparently, actual PBMs have a stricter definition of what is an appropriate reason for a prisoner being granted parole. Another notable difference was the factor of insufficient case detail. Actual PBMs listed this factor quite frequently when denying parole whereas mock PBMs listed this very rarely. Apparently, actual PBMs seek more information than do mock PBMs – and they err on the side of caution by denying parole. Actual PBMs


510


must be sensitized to the amount and type of information they require before making release decisions. This factor also surfaced in Carroll and Payne’s (1977) examination of the differences between parole decision makers and students’ perceptions of offenders. In their study, some experts refused to participate in the study because they felt uncomfortable with making a decision with so little information. Mock PBMs also differed from actual PBMs in the importance they placed on the prisoner obtaining an education. Mock PBMs listed failure to obtain an education as the third most common factor for denying parole whereas actual PBMs listed it less frequently. The fact that the mock PBMs were all college students makes them likely to value education. College students likely believe that obtaining an education is the key to rehabilitation and lessening the likelihood of recidivating. Another difference is the prisoner’s desire to improve his life which was considered more frequently by mock PBMs than actual PBMs. When considering parole, 20.1% of mock PBMs, whereas only 3.7% of actual PBMs listed the life improvement factor as important to their decision. This difference may be explained by attribution theory. Attribution is the process humans employ to understand what caused a behavior (Moskowitz, 2005). Attribution theory proposes that humans have a tendency to describe behavior as if it were intended or caused by the personality of the actor rather than by the actor’s situation (Moskowitz, 2005). Mock PBMs may attribute the cause of desiring to improve one’s life to dispositional rather than situational factors. This indicates that mock PBMs may assume that the prisoner had personal control over his ability to control his life, while actual PBMs assume that the prisoner has little control over improving his life because of the strict demands of the prison system. Thus mock, but not actual, PBMs listed

this factor. This is similar to Carroll and Payne’s (1977) study which suggested college students interpreted case summary information of a prisoner by making personal attributions, whereas actual PBMs did so less often. Another interesting difference concerned disagreement as to whether certain factors should be a reason to grant as opposed to deny parole. For instance, mock PBMs were divided as to whether the prisoner’s behavior and involvement in programming was positive (influencing a decision to grant parole) or negative (influencing a decision to deny parole), whereas actual PBMs were less divided – most considered this negative because there was not enough of it. Perhaps actual PBM have experiences that suggest what normally expected behavior is for a prisoner, whereas mock PBMs have no such experience to draw upon. It is also interesting that the second most common factor cited by mock PBMs for granting was that the prisoner deserved the benefit of the doubt. In contrast, none of the actual PBMs cited this factor. Perhaps mock PBMs give the benefit of the doubt to prisoners, such as is socially normative in the real world. In contrast, actual PBMs separate real world norms from norms required of PBMs. Specifically, PBMs are more acutely aware of the risks of giving second chances (e.g., the prisoner recidivates), and are not likely to list this factor. Actual and mock PBMs also differ in frequency of listing the interpersonal other factor. When describing violent behavior or positive behavior, several mock PBMs discussed the behaviors in relation to other prisoners whereas actual PBMs never directly listed others being affected. Mock PBMs may be more sensitized to prisoner behavior, imagining real people being affected (e.g., another prisoner is injured) by the prisoner. Actual PBMs may recognize that fights, for example, are common


Actual and Mock Parole Board Members place and may not associate prison behaviors with affected others, instead considering it a general disruption to the prison’s environment. Actual and mock PBMs also differ in how the prisoner’s inconsistent behaviors affect their decisions. Consistency was listed as the second most mentioned factor for denying parole by mock PBMs and the eighth most mentioned factor by actual PBMs. This difference may be explained by attribution theory. Mock PBMs may be more sensitized to inconsistent prisoner behaviors than actual PBMs because mock PBMs may perceive the prisoner as trying to fake positive change to achieve parole (e.g., verbally resolving to be peaceful and getting into a fight later) and thus judge the prisoner’s disposition to be the cause. Actual PBMs may recognize more clearly than mock PBMs that prison is a violent environment and commitments to change may coexist with intermittent displays of violence that may be outside the control of the prisoner. Actual and mock PBMs also differed in the complexity of their decisions. Mock PBMs listed more factors that affected their release decisions than actual PBMs. This may be explained by actual PBMs using heuristics because of their training as experts. Heuristics are cognitive shortcuts individuals make to simplify decision-making (Tversky & Kahneman, 1973). As trained experts, actual PBMs may have been taught some factors are important and others are irrelevant (e.g., giving the prisoner the benefit of the doubt). Mock PBMs, as non-experts, have had no such training and may consider extraneous information important. This is similar to research on experts vs. non-experts which found that courtroom judges (Ebbesen & Konecni, 1975) make decisions using fewer factors than jurors (Schwarz et al., 1991). These one or two important factors can serve as heuristics – shortcuts in decisionmaking. Like PBMs, judges often have

511

more exposure to judicial decisions than community members and may realize that heuristics can be used to make quick and (usually) reliable decisions. It is possible that the expertise of PBMs may be the reason why they considered fewer factors to make their decisions. In contrast, mock jurors have no experience and have not developed a sense of which factors should or should not be considered; hence they list more. Factors influencing actual PBM release decisions were consistent with findings from other research on PBM decisionmaking, while factors influencing mock PBMs were not. Caplan (2007) identified six common factors most prevalent across empirical reviews for how PBMs determine parole: (1) institutional behavior, (2) crime severity, (3) criminal history, (4) incarceration length, (5) victim input, and (6) mental illness. The current research generally confirms these factors as being salient factors considered by actual and mock 2 PBMs who denied parole. However, this research suggests Caplan’s findings do not apply well to factors considered important to mock PBMs for granting parole. In sum, the current study found a few similarities, but mostly differences between actual and mock PBMs decision-making. Research Question Three. Research Question Three concerns whether actual and mock PBMs differ in perceptions of a prisoner. Actual PBMs perceived the prisoner as more likeable and credible than mock PBMs. This may be because actual PBMs are involved regularly with aggressive prisoners who have committed crimes such as murder and are habituated to violence by aggressive prisoners. Actual PBMs may perceive violence in prison to be situational, part of the game of surviving prison, and may perceive someone who has committed murder as just as likely to behave violently in prison as other prisoners. This is supported by actual PBMs who never cited other prisoners being affected


512


by the prisoner’s violent behavior. Carroll (1978) found that PBMs made very few internal attributions toward murderers because they knew that most murders were caused by situational events and were not predictive of future violence. Mock PBMs may be more sensitized to violence committed by someone who has already committed murder, perhaps labeling the prisoner as a deviant – a ‘‘murderer.’’ The sensitization of mock PBMs to the prisoner’s violence is demonstrated when they listed others as being affected by the prisoner’s violent behavior quite often. This could influence the mock PBMs to perceive the prisoner as less likeable and credible than actual PBMs. It is interesting that perceptions of the prisoner and release decisions seemed to have opposing patterns. Mock PBMs were more likely to release the prisoner, but had more negative perceptions of the prisoner. This may indicate that these are two different kinds of decisions. How likeable or credible the prisoner is may have little to do with whether a prisoner should be released. Furthermore, accounting for perceptions of the prisoner played an important role in the findings because the difference between mock and actual PBMs’ release decisions was only marginally significant unless perceptions were controlled; when perceptions were controlled actual PBMs made significantly more punitive decisions than mock PBMs. Implications for Expert versus Non-Expert Research The common theme in all three Research Questions is that many of the differences could be explained by the relative experiences and background information of mock and actual PBMs. Students and PBMs have different socially constructed world views, based on their different experiences and knowledge, which might affect their attitudes, beliefs, expectations

and attributions. Research supports the notion that background knowledge and experience can affect perceptions. Psychology students and engineering students read about an accident that was a result of both technical errors and human errors (Svenson, Lekberg, & Johansson, 1999). The engineering students were better able to identify the technical errors, while psychology students were better able to identify the human errors, suggesting that this was a result of greater knowledge and involvement in their respective fields of study. In essence, individuals in different fields draw on different experiences, which lead them to make different decisions regarding the cause of the accident. This could certainly be the case with our sample; mock and actual PBMs have differing experiences, which may lead them to make different decisions based on different factors and decision-making processes. While much research (Shanteau & Gaeth, 1981) has found that experts and non-experts often come to similar decisions, much of this research is based on non-social decisions (such as deciding which corn is best). Social decisions (e.g., parole decisions) are likely more complex and thus may lead to more variability and differences between experts and non-experts. Human behavior is complex, difficult to predict, and can have serious consequences. Thus it may be a different kind of decision than deciding which grain will be more successful. The issues surrounding decision-type and previous experiences highlight the complexity of studying the differences between experts and nonexperts. Implications for Community Sentiment The community has perceived that parole board decisions are too lenient (Petersilia, 2003). But in fact, the community sample in this study was more lenient than the PBMs. This may be explained by


Actual and Mock Parole Board Members misperceptions of the public. Roberts and Stalans (1997) note that the public’s perceptions of parole are often founded on inaccurate information and this may explain why public perceptions are sometimes punitive. That is, the public often misperceives how frequent and to whom parole is granted. Roberts and Stalans suggest that it is possible that correcting the public’s misperception may result in increased community support for discriminatory parole. To date, no research (of which we are aware) has yet tested how to correct this misperception and whether a correction would change public perceptions and beliefs about parole. It is important that future research test how to disabuse the public’s misperceptions of leniency and parole decisions. This is important because the publics’ beliefs influence their perceptions of discretionary release as legitimate and PBMs as legitimate authorities. If community perceptions can be corrected, shifting community sentiment, the public will develop higher regard for the purpose of discretionary release. Communities will benefit when parole is denied to prisoners still at risk to reoffend; and prisons will benefit because discretionary release helps reduce prison overcrowding by releasing those seen as unlikely to reoffend. This study also has implications for how paroling authorities should consider determining whether a prisoner qualifies for parole. Several of the factors listed by community members as important to their release decision were never listed by PBMs. One example of a factor commonly mentioned by community members when parole was denied was the inconsistency of the prisoner’s behavior. Paroling authorities may benefit from adding a factor such as this, which is valued by the public, to its recommended guidelines for PBMs to consider. By selectively implementing factors considered important to the public, paroling authorities can increase their

513

perceived legitimacy. This assumes that the public is knowledgeable about the parole process. Implications for Using Students as Proxies Jury research shows that mock juror decisions using college students are generally equivalent to actual juror decisions, but when differences are found, students tend to be more lenient (Bornstein, 1999). The current research comports with this general rule. From a research standpoint, it would be ideal to conclude that mock and actual PBM decisions are identical and thus future studies could use students as mock PBMs. However, this research finds that students may make different decisions – and make those decisions differently – from actual PBMs. This research supports the notion that college students are truly non-experts in the domain of parole decision-making, while PBMs are experts who have had extensive training and practical experience applying the guidelines. If training and experience can account for the differences found in this study, then expertise does matter and students are not adequate proxies for PBMs. However, this finding does not necessarily translate into students’ inability to proxy for other types of legal decisions such as those made by juries. Perhaps, decision-making by mock and actual jurors is quite similar because no training is required to serve as a juror. Thus, just because students are not sufficient proxies for PBMs, they may still be useful in lawpsychology research as a whole. Limitations and Future Research One of the limitations of this research was that the response rate from actual PBMs was quite low (23%). PBMs who responded to this study may be different from PBMs who did not respond. PBMs who responded may be possibly more


514


confident, or more consistent in their release decisions. Confident PBMs may use fewer stereotypes or personal biases in their judgments and feel less vulnerable to researcher’s criticism of the past (e.g., inconsistency). This is also a limitation because it limits the amount of data available; for instance, the factors affecting release decisions were derived from a small number of actual PBMs. Actual PBMs may be guarded about their decisionmaking because of negative attention from the past which they received from research and the media portraying them in a negative light (Petersilia, 2003). To address this limitation, future research should develop relationships with PBMs through parole board organizations (e.g., Association of Paroling Authorities International) and address any other concerns that PBMs may have with research participation. A second limitation of this study is that mock PBMs were students from an introductory criminal justice course, not a random sample of community members. While this sampling helps determine whether these students may be good proxies for actual PBMs, it also limits the conclusions that can be made about general community sentiment. Students may be different from other community members who are unfamiliar with parole release procedures or less educated about the justice system. Further, students may have learned information from their classes that could have influenced their parole decisions. Thus, to conclude that actual PBMs are more punitive than mock PBMs may be limited because mock PBMs were not randomly selected from community members. To address this limitation future research should determine whether a more representative sample from the community acting as mock PBMs differ from college students acting as mock PBMs. A third limitation is the short length of the prisoner’s case summary. The case

summary was 650 words, whereas a real case summary is much longer. Actual PBMs work with more detailed information covering many aspects of the prisoner. PBMs pointed this out: 25% of the actual PBMs mentioned they did not have enough information to make a decision. The case summary may have also been limited because the prisoner and summary were fictitious. The study lacked the impact of a real life decision determining the fate of another human being and all the lives the prisoner could affect in the community. Theoretically, it may be more likely that participants would put more thought into their decision and decision process if the participants were to assess a real prisoner. This lack of verisimilitude is a similar issue in jury decision-making studies that use written trial summaries. To address this limitation future research should seek to make the decision-making materials and environment more believable and information rich. While more exhaustive case summaries would obviously take more time for participants to read, they may help participants to make more informed and confident decisions. This presents a natural challenge because the lengthier the research experience the less likely participants are to participate without increasing compensation. The case summary also only described a prisoner who committed one type of crime. Other research suggests that mock and actual PBMs may differ in their decisions when different types of crimes are detailed (Carroll & Payne, 1977). Finally, a fourth limitation of this research is that this research does not clarify what drives the effects that were found. Differences between actual and mock PBMs could have been influenced by many factors such as the age or gender of the PBMs (actual PBMs included more males – 67% – than mock PBMs – 48%). This study suggests that most differences can be explained by the different experiences and backgrounds of the two groups.


Actual and Mock Parole Board Members This proposal needs clarification and support from future research. If our results are primarily a result of expertise, future studies could examine whether college students can be trained to be experts in parole decision-making. Studies could attempt to determine how much training is required for students to learn to make decisions similar to PBMs. Once it is understood how much and what type of training is required to make PBM decisions, college students could be effectively trained and represent adequate proxies for PBMs.

Acknowledgements This research was supported in part by the Office of Undergraduate Research at UNR. We thank Stacy Shamblin for her assistance in conducting this study. Portions of this research were presented at the 2009 American Psychology-Law Society conference.

Notes 1.

Conclusion Parole decision-making research is critical to implementing best practices that ensure the safety of communities while balancing the needs of increasingly overcrowded prison systems. However, it is difficult to study actual PBMs, and the current study indicates that college students are not necessarily the best proxies. Thus, challenges of studying parole decision-making remain. The prison systems have adjusted their release procedures to respond to the demands of the public, however, this might not be optimal if the community is not well informed as to how PBMs actually make decisions. Research such as this continues to challenge public misperception; for instance there is a misperception that PBMs are more lenient than the public prefers. Thus it is important to educate the public as to the realities of parole board decision-making to ensure that perceptions of the legitimacy of the parole system do not suffer. A better educated public can express community sentiment that legitimizes the authority of PBMs in future parole decisions. This research suggests state policy-makers should reconsider the importance of discretionary release, perhaps in favor of some combination of mandatory and discretionary release.

515

2.

This study was originally designed to assess differences between actual and mock PBMs when making release decisions about a prisoner who experienced a religious conversion. Three conversion conditions and a control were used. Because of the sparsity of responses by actual PBMs (36 actual PBMs spread approximately equally across four conditions) the research design was altered and the effect of religious conversion was not evaluated for this study. For the purposes of this study, all actual PBM responses were grouped together and compared collectively to all mock PBM responses. Only four of Caplan’s (2007) six factors were overtly mentioned in the case summary (mental illness and victim input were excluded) provided to the participants. This explains the low frequency of the two factors in Table 1.

References Association of Paroling Authorities International. (2008). 2007-2008 survey report. Retrieved July 27, 2009, from http://www. apaintl.org/en/aw_surveys.html Bornstein, B.H. (1999). The ecological validity of jury simulations: Is the jury still out? Law and Human Behavior, 23, 75–91. doi: 10.1023/A:1022326807441 Burke, P.B. (Ed.). (2003). A handbook for new parole board members. Association of Parole Authorities. Retrieved July 27, 2009, from http://www.apaintl.org/en/aw_publica tions.html Caplan, J.M. (2007). What factors affect parole: A review of empirical research. Federal Probation, 71(1), 16–19. Carroll, J.S. (1978). Causal attributions in expert parole decisions. Journal of Personality and Social Psychology, 36, 1501–1511. doi: 10.1037/0022-3514.36.12.1501


516


Carroll, J.S., & Payne, J.W. (1977). Crime seriousness, recidivism risk, and causal attributions in judgments of prison term by students and experts. Journal of Applied Psychology, 62, 595–602. doi: 10.1037/00219010.62.5.595 Cumberland, J., & Zamble, E. (1992). General and specific measures of attitudes toward early release of criminal offenders. Canadian Journal of Behavioural Science, 24, 442–455. doi: 10.1037/h0078752 Ebbesen, E.B., & Konecni, V.J. (1975). Decision making and information integration in the courts: The setting of bail. Journal of Personality and Social Psychology, 32, 805– 821. doi: 10.1037/0022-3514.32.5.805 Finkel, N. (1995). Common sense justice: Jurors’ notion of the law. Cambridge, MA: Harvard University Press. Gallup Poll News Service. (1937, February). The Gallup poll #69. Retrieved July 27, 2009, from http://institution.gallup.com Hughes, T., Wilson, D.J., & Back, A.J. (2001). Trends in state parole, 1990-2000. Washington, DC: Bureau of Justice Statistics. Kahneman, D., & Tversky, A. (1973). On the psychology of prediction. Psychological Review, 80, 237–251. doi: 10.1037/h0034747 Kalven, H., & Zeisel, H. (1966). The American Jury. Boston: Little, Brown, and Company. Kardes, F.R., Muthukrishhnan, A.V., & Pashkevich, V. (2005). On the conditions under which experience and motivation accentuate bias in intuitive judgment. In T. Betsch & S. Haberstroh (Eds.), The routines of decision making (pp. 139–156). Mahwah, NJ: Lawrence Erlbaum. Maguire, K., & Pastore, A. (Eds.). (1995). Sourcebook of criminal justice statistics, 1994. Washington, DC: U.S. Department of Justice, Bureau of Justice Statistics. Montana Board of Pardons. (2009). Administrative rule 20.25.102 board training. Retrieved August 7, 2009, from http:// mt.gov/bopp/adminrules/Admin_rule_20.25. 102.asp Moskowitz, G.B. (2005). Social cognition. New York: The Guilford Press. Petersilia, J. (2003). When prisoners come home: Parole and prisoner reentry. New York: Oxford. Phillips, J.K., Klein, G., & Sieck, W.R. (2004). Expertise in judgment and decision making: A case for training intuitive decision skills. In D. Koehler & N. Harvey (Eds.), Blackwell handbook of judgment and decision making (pp. 297–315). Oxford: Willey-Blackwell.

Roberts, J., & Stalans, L. (1997). Public opinion, crime, and criminal justice. Boulder, CO: Westview Press. Schwarz, N., Strack, F., Hilton, D., & Naderer, G. (1991). Base rates, representativeness, and the logic of conversation: The contextual relevance of ‘‘irrelevant’’ information. Social Cognition, 9(1), 67–84. Shanteau, J. (1992). How much information does an expert use? Is it relevant? Acta Psychologia, 81, 75–86. Shanteau, J., & Gaeth, G.J. (1981). Evaluation of the field method of soil texture classification: A psychological analysis of accuracy and consistency (Tech. Rep. No. 79-l). Manhattan, KS: Kansas State University, Department of Psychology. Shanteau, J. , & Stewart, T.R. (1992). Why study expert decision making? Some historical perspectives and comments. Organizational Behavior and Human Decision Processes, 53, 95–106. doi: 10.1016/0749-5978(92)90057-E Stivers Ireland, C., & Prause, J. (2005). Discretionary parole release: Length of imprisonment, percent of sentence served, and recidivism. Journal of Crime and Justice, 28, 27–49. Svenson, O., Lekberg, A., & Johansson, A.E.L. (1999). On perspective, expertise and differences in accident analyses: Arguments for a multidisciplinary integrated approach. Ergonomics, 42, 1561–1571. Tonry, M.H. (Ed.). (2000). The handbook of crime and punishment. New York: Oxford University Press. Trumbo, D., Adams, C., Milner, M., & Schipper, L. (1962). Reliability and accuracy in the inspection of hard red winter wheat. Cereal Science Today, 7, 62–71. Tversky, A., & Kahneman, D. (1973). Availability: A heuristic for judging frequency and probability. Cognitive Psychology, 5, 207–232. doi: 10.1016/0010-0285(73)90033-9 Tyler, T. (2006). Restorative justice and procedural justice: Dealing with rule breaking. Journal of Social Issues, 62, 307–326. United States Parole Commission. (2003a). History of the federal parole system. Washington, DC: U.S. Department of Justice. United States Parole Commission. (2003b). Rules and procedures manual. Chevy Chase, MD: U.S. Parole Commission. Vermont Parole Board. (2007). FY 2007 annual report. Retrieved from August 7, 2009, from http://doc.vermont.gov/about/paroleboard/pb-annual-report/view