Abstract
Research suggests that a defendant's history of experiencing childhood abuse, and its effects on their life and later decision-making, may impact public support for a defendant's sentencing, particularly mitigation. However, no existing research has examined how and why sentencing support may vary based on the time period when the abuse occurs during a defendant's childhood. This experiment, using a sample of the U.S. public (N = 400), examines how the age at which a defendant's childhood physical abuse occurs affects lay support for the goals of their sentencing. We hypothesized that participants with higher levels of social and biological trait essentialism would moderate their increased support for more punitive sentencing goals—particularly when a defendant was abused earlier, rather than later, in childhood. Results suggest that social essentialism is associated with increased support for restoration and rehabilitation toward defendants with histories of childhood physical abuse, potentially indicating that the public views the effects of child abuse as more of a social, versus biological, process which may affect support for utilitarian punishment goals.
1 INTRODUCTION
Although there is evidence to suggest that a person's history of experiencing abuse during childhood can increase the likelihood of exhibiting future criminal behavior (Currie & Tekin, 2006; Fox et al., 2015; Weiler & Widom, 1996; Widom, 1989), childhood abuse in a defendant's background is still only rarely viewed as a mitigating factor to their sentencing outcomes in U.S. courts (Berryessa, 2021). This has led to a burgeoning body of literature arguing that a defendant's childhood abuse should be considered a meaningful mitigating factor to punishments across types of cases and courts (Bagaric et al., 2019; Nunez et al., 2007; Stevenson et al., 2010).
There has also been some limited research specifically on the public's views of and support for whether a defendant's history of childhood abuse should be considered as mitigating to their sentencing (Bell Holleran et al., 2016; Platania & Konstantopoulou, 2014). Gauging the public's views on whether a defendant's childhood abuse should be weighed in different sentencing contexts has been considered an important area of work; scholars and even the U.S. Sentencing Commission have argued that sentencing guidelines should engage the views of the public and factors they believe are important to consider in order to better ensure just, fair, and legitimate outcomes from our sentencing systems (Berryessa, 2021; Maxfield et al., 1996; Roberts & de Keijser, 2014). Although some evidence points toward public support for considering a defendant's childhood abuse in sentencing contexts and that it may suggest the potential need for more rehabilitative punishments (Berryessa, 2021; Platania & Konstantopoulou, 2014), other literature remains mixed (Najdowski et al., 2009; Stevenson et al., 2010).
At the same time, different ages during childhood come with differences in cognitive development, self-control, and ideas of responsibility—with views of the public on these constructs also potentially impacting how a defendant's childhood abuse and victimization may be considered in punishment contexts (Fogel et al., 2023). Indeed, there are no existing studies that have assessed whether the age at which a defendant experienced that abuse may affect the public's views on and support for their sentencing.
Further, the age at which a defendant experienced childhood abuse and whether it should be considered mitigating to sentencing could also vary based on other types of psychological beliefs. Particularly, thinking stemming from trait essentialism, which is defined as people's general views or beliefs that concepts, items, or individuals have particular inherent characteristics that shape what or who they are (Berryessa, 2020; Haslam et al., 2000, 2002, 2006). This is an unconscious process that can lead to stereotyping and prejudice of a person and their outcomes based on when a characteristic or behavior is believed to have become a part of a person's identity (Bastian & Haslam, 2006; Gowaty, 2018). Trait essentialism encompasses two areas or sets of beliefs: social essentialism and biological essentialism. Social essentialism is the belief that social backgrounds are responsible for behavior formation, while biological essentialism emphasizes the role of biology and pre-birth factors in trait development (Rhodes & Mandalaywala, 2017; Soylu Yalcinkaya et al., 2017). If childhood abuse occurs earlier or later in life, such beliefs could impact the views of the defendant, their responsibility, and ultimately support for sentencing outcomes. Thus, the current research experimentally examines how the age at which a criminal defendant experienced childhood abuse may affect public support for the goals of sentencing and whether this support may be moderated by their levels of social and biological trait essentialism.
1.1 Childhood abuse and criminality
A person's history of childhood abuse has been known to be correlated with chronic issues in adulthood, including post-traumatic stress disorder, emotional distress, physical health problems, and problems with future relationships (Chang et al., 2019; Kendall-Tackett, 2001; Strathearn et al., 2020). Further, there is a myriad of evidence and criminological theories to suggest that sexual, physical, and/or emotional abuse during childhood can increase the likelihood of an individual's criminal offending as an adult (Herrenkohl et al., 2022; Lansford et al., 2007; Widom, 1989). According to Social Control Theory, abuse can disrupt the social bonds between a child and parent that are needed for prosocial behavior and increase the risk of later offending (Hirschi, 1998; Widom, 2017). Social Learning Theory argues that childhood abuse victims learn and then mimic the behaviors they observe and experience—such as using violence or manipulation to achieve goals or resolve conflict—and may later lead to offending (Widom, 2017). Life Course Theories posit that life-course persistent offenders have neuropsychological deficits before or soon after birth, which can be the result of child abuse, and these deficits coupled with a poor social environment may lead to the highest risk of a lifetime of criminal behavior (Brennan & Raine, 1997; Moffitt, 1993). Finally, General Strain Theory states that abuse leads to high levels of acute stress, which can then affect development and lead to changes in the ways that abuse victims respond to environmental stimuli (Agnew, 2001; Currie & Tekin, 2012). Evidence shows that childhood abuse strongly predicts later recidivism as an adult, even after controlling for peer influences, academic performance, and mental health or substance abuse issues (Kim et al., 2016).
Abuse that occurs at different periods of a child's life can also affect their development and later criminal behavior. Abuse commonly occurs across all childhood age ranges, including before birth, and affects all ages under 18 (Guastaferro & Shipe, 2023; NCANDS, 2020). The prevalence of child maltreatment across age groups has recently been captured from national data: 10% were 15 years or older, 19% were 11–14 years old, 25% were 6–10 years old, 17% were three to 5 years old, and 28% were under two years of age and prenatally (Guastaferro & Shipe, 2023). Prenatal development is crucial for proper physical functioning, such as cognitive, visual, and auditory functions, and emotional and sensory development (Berksu et al., 2021); as such, abuse prenatally (e.g., exposure to alcohol or drugs in utero or violence inflicted on the mother that affects the fetus) can affect a fetus's brain development—leading to emotional, physical, and intellectual issues after birth (Hiscox et al., 2023).
In early to mid-childhood, multiple stressful events in early childhood, including abuse, can affect emotional regulation, reduce cognitive abilities, and contribute to poor stress reactions (Fogel et al., 2023). Indeed, people develop more strength, thinking and reasoning abilities, emotional management and social skills, and language and fine motor skills at two to 5 years old (Alberta Health Services, 2002), while developmental milestones from eight to 11 years old include physically developing due to puberty, learning how to cope with social pressures and heightened emotional sensitivity, and forming more empathy and understanding (Weinstein, 2022). Finally, at ages 14–17, individuals should have mostly finished maturing physically, developed more independence, and be capable of reasoning through their choices about what is right or wrong (CDC, 2019a). Yet individuals who experience abuse later in adolescence and closer to adulthood may also suffer emotional, psychological, and other developmental effects but could be perceived as further removed from the impact of the abuse and having access to more resources to alleviate its harmful influences (Fogel et al., 2023). In contrast, those experiencing earlier abuse may be seen as not yet having the independence or ability to avoid such exposure (Fogel et al., 2023).
1.2 Childhood abuse and sentencing
During sentencing, mitigating and aggravating factors can be taken into consideration, with mitigating factors contributing to support for a lesser or reduced penalty and aggravating factors contributing to support for a higher or harsher punishment (Macrae, 2022). Despite decades of evidence suggesting the criminological effects of childhood abuse on development and later offending, whether a defendant's childhood abuse is viewed as mitigating to their criminal sentencing outcomes remains inconsistent (Berryessa, 2021). Historically, a defendant's childhood abuse has been more likely to be presented during and viewed as mitigating in capital cases, which are cases in which the death penalty is faced (Berryessa, 2021; Denno, 2019). The U.S. Supreme Court has stated that juries should consider a defendant's childhood abuse, particularly linked with abandonment and substance abuse of family members (Williams v. Taylor, 2000) and foster care involvement and neglect (Wiggins v. Smith, 2003), as potential mitigating factors to weigh during capital proceedings. Additionally, the likelihood of its consideration during sentencing in non-capital cases has been significantly mixed at both the federal and state levels (Berryessa, 2021; Denno, 2019).
However, childhood abuse does not appear as a mitigating factor in any state sentencing guidelines, and some explicitly disallow it to be considered during sentencing (Baron-Evans & Hofer, 2010). In states that do allow childhood abuse to be considered, it is often presented and weighed in highly variable ways, as well as weighed against other case and defendant-specific factors related to a defendant's offense (Bagaric et al., 2019). As childhood abuse contributes to emotional and physical problems that highlight disadvantages in functioning between those who have been and have not been, many legal scholars and judges have considered it as a potential deficits-based mitigating factor—similar to disabilities, mental illness, intellectual deficits, and trauma—for sentencing outcomes in non-capital settings (Bagaric et al., 2019; Baron-Evans & Hofer, 2010; Vartkessian, 2019).
Regardless of its current use in courts, many legal scholars have argued that childhood abuse should normatively and philosophically mitigate the goals of sentencing pursued in federal and state sentencing guidelines (Bagaric et al., 2019). Gilman (2012) argues that a defendant's childhood abuse has similarities to other accepted legal excuses, like duress, and other accepted mitigating factors, like severe emotional distress—suggesting their life choices, or their ability to “choose otherwise,” are in some way disturbed by their past circumstances that they have experienced (Berryessa, 2021; Najdowski et al., 2009; Piazza et al., 2013; Stevenson et al., 2010). Moreover, as childhood abuse can also affect development related to later decision-making and judgment (Widom, 2017), a defendant's capacity to make decisions and corresponding criminal responsibility may also be considered compromised (Bagaric et al., 2019).
While mitigation is traditionally related to incarceration-based sentences, the goals of sentencing policies, systems, and decisions in the U.S. traditionally encompass five purposes: rehabilitation, incapacitation, restoration, deterrence, and retribution (see Maxfield et al., 1996; Sample, 2018 for definitions). These goals are commonly partitioned into those that speak to “just deserts” theory (punishment is proportional to the moral wrong committed) or utilitarianism (punishment should be focused on reducing the likelihood of future wrongdoing) (Carlsmith et al., 2002; Weiner et al., 1997). In cases involving a defendant's history of experiencing childhood abuse, such evidence is likely relevant to the range of sentencing goals within these two philosophies (Litton, 2005; Tetterton & Brodsky, 2007). Berryessa (2021) suggests that rehabilitation in cases involving defendants with histories of childhood abuse could help them manage and treat the effects of their experiences, take back “responsibility” for their actions, and eventually reduce the likelihood of their future offending. Therefore, in addition to incarceration-based sentences, considering a defendant's childhood abuse is likely relevant to sentence contexts related to broader goals or theories of punishment (Berryessa, 2021).
1.3 The public's views of a defendant's childhood abuse in sentencing contexts
Although judges in court pass down the majority of sentencing decisions in the U.S. (American Bar Association, 2019), public sentiments remain highly important to the state's moral authority to punish and the development of sentencing guidelines (Zimring & Johnson, 2006). Following the Sentencing Reform Act of 1974, the U.S. Sentencing Commission was pressed to consider community views when creating or amending sentencing guidelines to complement public beliefs on what are just and appropriate punishments across crimes and circumstances (U.S. Sentencing Commission, 1995). Indeed, gauging the public's views and support for different goals of sentencing, as well as when and in what contexts mitigation is warranted, are crucial in order to better ensure that capital and non-capital sentencing practices are viewed as fair and legitimate (Maxfield et al., 1996).
Further, data surrounding whether members of the public support mitigation in sentences for different cases, defendants, and circumstances have also been emphasized by the democratizing punishment movement, which considers sentencing and court-enforced punishments as a collective social action made on behalf of the community; as such, the public should help to determine the nature, extent, and circumstances of the state punishments (Roberts & de Keijser, 2014; Robinson, 2017). Public opinion can also aid and permit the development and passage of criminal justice reform, including sentencing reform, as an essential tool to identify a consensus for policy (Drakulich, 2022; Johnson & Huff, 1987; Vera Institute, 2018). Another reason to consider public sentiments regarding sentencing guidelines relates to theories surrounding procedural justice, which state that peoples' acceptance and endorsement of the legal system and authorities depend on their perceptions of the fairness of its procedures and processes during which legal decisions and rules are made and applied (Tyler, 2003; Tyler & Lind, 2001). Examining public support for sentencing guidelines ensures that legal policies are in some way shaped by their opinions, which in turn contributes to their perceptions of a system that practices fair and just sentencing policies as well as better overall trust and positive views of the legal system as a whole and its authorities (Sunshine & Tyler, 2003; Tyler & Lind, 2001).
Recent studies have begun to examine the public's consideration of a defendant's childhood abuse in relation to mitigation and sentencing outcomes in different types of cases. In several studies, mock jury members found childhood abuse, both physical and sexual, in a defendant's background mitigating in death penalty cases (Bell Holleran et al., 2016; Platania & Konstantopoulou, 2014). However, in a cross-sectional survey study involving a capital case, mock jurors found a defendant's childhood physical abuse as aggravating and as a rationalization that the abuse had led to permanent damage and an inability to be rehabilitated (Stevenson et al., 2010). Experiments involving non-capital assault cases have found varying effects. One experimental vignette study found that a defendant's childhood sexual abuse increased public support for rehabilitative sentences and decreased support for sentences involving incapacitation (Berryessa, 2021). However, another study consisting of two mock trial experiments in which the defendant was abused physically and sexually, as well as neglected, reported decreased support for a defendant's responsibility and their likelihood of responding to rehabilitation when compared to those who were not abused (Najdowski et al., 2009). Stevenson (2009) conducted a review which found that a juvenile defendant's childhood abuse was most often viewed by the public as mitigating to prison time across a sample of mock juror studies, but that juvenile court officials often viewed it as an aggravating factor to sentencing.
To date, existing research has not yet assessed whether the age at which a defendant experienced that abuse may affect the public's support for sentencing. Indeed, specific stages of childhood can differentially impact cognitive development, self-control, and other characteristics that can affect decision-making (Fogel et al., 2023). Indeed, Griffin and Sallen (2013) argue that the age at which a person has experienced abuse in childhood at different stages of biological and social development, as well as individual views of what it means to be a child, could influence lay views of how a defendant's childhood abuse should be considered during sentencing; for example, a defendant experiencing abuse earlier in childhood, such as before or soon after birth, could be perceived as resulting in behavior or traits that are more unchangeable or dangerous—as the “damage has been done”—which could increase support for more punitive sentencing goals and decrease support for a defendant's rehabilitation and restoration.
Likewise, individual-level beliefs on when and how a person becomes “who they are,” particularly during childhood, may also bear on how members of the public view the purposes of sentencing—especially if “who they are” is considered inherent and unchangeable (Xu et al., 2022). If a defendant's childhood abuse occurs earlier or later in life, individual-level beliefs on when and how a person becomes “who they are,” particularly during childhood—like trait essentialism—could moderate public views of the defendant, their responsibility, and ultimately support for sentencing outcomes (Berryessa, 2020; Griffin & Sallen, 2013). People who show high levels of trait essentialism typically believe that other people's personality, behavior, and background traits are innate, determined, often immutable, and that people who share the same characteristics (or belong to the same “category”) are inherently alike (Gelman, 2003; Ryazanov & Christenfeld, 2018). Such views, based on the uniformity, immutability, discreteness, exclusivity, and biological basis of a person and their characteristics (see Haslam et al., 2000) and, as mentioned above, consider two sources of an innate characteristic: social (social backgrounds are responsible for the way that an individual's “true essence” and behavior is formed) and biological (emphasizing the role of biology, genetics, and other factors pre-birth in behavior forming and trait development) (Rhodes & Mandalaywala, 2017; Soylu Yalcinkaya et al., 2017).
As essentialist thinking leads individuals to reduce others to specific attributes instead of viewing them as whole people (Berryessa, 2020), trait essentialist beliefs rely on using innate characteristics to explain social stereotypes and correspondingly may lead to prejudice and discrimination, including in punishment contexts (Haslam et al., 2006). This is why both categories (social and biological essentialism) of trait essentialism can lead to ideas of punitiveness and discrimination (Berryessa, 2020). If criminal acts are thought to be connected to a defendant's predetermined “immoral essence,” those who hold such views may support sentencing in ways that stem from societal stigmas and prejudices toward people with certain attributes (Xu et al., 2022). Most commonly, essentialist beliefs have been found to increase public support for harsher and more retributive punishments for defendants with certain attributes, such as mental illness or histories of sexual offending (Berryessa, 2020)—mainly via increased perceptions that certain defendants are more dangerous and more responsible for their crimes due to a “criminal essence” that is immutable to change (de Vel-Palumbo et al., 2018; Kraus & Keltner, 2013; Martin & Heiphetz, 2021; Newman et al., 2015; Xu et al., 2022). However, some research has shown opposite effects and that essentialist beliefs, particularly related to biological essentialism, can sometimes reduce moral condemnation for one's actions and reduce support for harsh sentencing outcomes(Dar-Nimrod & Heine, 2011; Martin & Heiphetz, 2021; Monterosso et al., 2005). Yet, as most commonly existing studies have found that biological essentialism predicts harsher support for or levels of punishment (Meyer et al., 2022), we suggest that may be reasonable to assume that this pattern would also be observed for this study as well.
Overall, lay people's levels of both social and biological trait essentialism could affect how they consider a defendant's history of childhood abuse in relation to mitigation and support for their sentencing outcomes. For example, social essentialist thinking could lead to beliefs that defendants with abuse in their childhood have been negatively affected by their social circumstances but could be amenable to treatment and benefit from more rehabilitative punishments if their social circumstances can be changed (Litton, 2005). On the other hand, higher levels of biological essentialist thinking could lead to views that abuse may have immutably disturbed a defendant's innate biological development and their future abilities to make good decisions—potentially suggesting they may not be amenable to rehabilitation and are likely to be a future threat (Fogel et al., 2023). This viewpoint could potentially lead to defendants who have experienced abuse early in life to be punished more punitively—due to views of their dangerousness and a desire to remove said danger from society (Fogel et al., 2023; Levitt, 2013).
Meanwhile, a defendant experiencing abuse later in childhood may be associated with more malleability and a higher potential for change for those who subscribe to stronger social or biological essentialist beliefs; indeed, changes or experiences during critical periods, such as prenatally or in early life, may be viewed as fixed, less malleable, and even inevitable (Weiss, 2016; Weiss & Diehl, 2021). Prior work suggests that characteristics believed to be “set” at birth, such as sex or race, are commonly viewed as more intrinsic or inherent to a person's sense of self than those acquired later in life (Bastian & Haslam, 2006).
1.4 Current study
The current study examines how the age at which a criminal defendant experienced abuse in their childhood affects public support for the goals of their sentencing in a non-capital criminal case. Participants' individual levels of social and biological trait essentialism are utilized as moderators to assess whether they affect the strength of the observed relationships between the defendant's age when they experienced childhood abuse and lay support for their sentencing.
Based on the literature above, we hypothesized that a defendant experiencing abuse earlier in childhood (before and soon after birth) would increase support for more punitive sentencing goals (retribution and incapacitation) and less support for rehabilitation and restoration. We also hypothesized that social essentialism and biological essentialism independently would positively moderate increased support for more punitive sentencing goals—particularly when a defendant was abused earlier in childhood. To clarify the moderation, we anticipated that defendants who were abused earlier would be significantly associated with support for more punitive punishment goals if social essentialist beliefs or biological essentialist beliefs were higher. Although there is mixed evidence for the relationship between increased punitiveness and higher levels of social essentialist beliefs, this hypothesis was formed based on the existing literature at large that has most commonly shown that trait essentialism can lead to more discriminatory and punitive support for sentencing practices.
2 METHOD
2.1 Participants
The primary data for this study was collected from a sample of U.S. adults obtained by Qualtrics Panel, a firm that collects responses from a pre-arranged pool of respondents who have agreed to be notified by a market research service to answer surveys. The research sample was collected based on quotas for race/ethnicity, age, sex, geographic location, and education that approximated the demographic makeup of the adult population in the U.S. (as compared to Census data). Respondents received around $2.00 once they completed the survey. Online panels have become more common in social science research because of their potential to reach and recruit samples that reflect the U.S. population (Chandler et al., 2019). Despite some concerns, research has found that online panels are most often reliable and a robust data source for social science experimental research (Mullinix et al., 2015; Walter et al., 2019; Weinberg et al., 2014).
An a priori power analysis was conducted, and for four groups, power of 0.99, alpha of 0.05, and a medium effect size, a sample size of 384 was calculated for sufficient power. The sample initially began with 445 respondents, but 45 were dropped due to incomplete survey responses. The 45 participants that were dropped included those who failed the attention check (and were prevented from completing the survey) or did not complete the survey fully. This resulted in a final analytic sample of 400. Models were run with and without the 45 exclusions, but the results were not significantly different. All research complied with the authors' Institutional Review Board. Respondents were required to provide their consent by agreeing to a digital informed consent document before starting the survey. The demographics of the sample, presented in full in Table 1, are similar to those of the U.S. population (U.S. Census Bureau, 2021). Data collected from this sample were also used in a separate inquiry by Silver and Berryessa (2023).
TABLE 1. Demographics (N = 400).
Age | |
18–24 years | 12% (N = 48) |
25–34 years | 18% (N = 72) |
35–44 years | 17% (N = 68) |
45–54 years | 18% (N = 72) |
55–64 years | 16% (N = 64) |
65+ years | 19% (N = 76) |
Gender | |
Male | 48.75% (N = 195) |
Female | 51.25% (N = 205) |
Education | |
High school graduate or less | 20.05% (N = 82) |
Some college | 21.75% (N = 87) |
Associate degree in college (2-year) | 13.5% (N = 54) |
Bachelor's degree in college (4-year) | 30% (N = 120) |
Master's degree | 11.5% (N = 46) |
Doctoral degree | 1% (N = 4) |
Professional degree (JD, MD) | 1.75% (N = 7) |
Race | |
White Non-Hispanic | 62% (N = 248) |
Black or African American | 12% (N = 48) |
Hispanic | 16.5% (N = 66) |
Asian | 5% (N = 20) |
Other | 4.5% (N = 18) |
Geographic region | |
South | 37% (N = 148) |
West | 23% (N = 92) |
Midwest | 22% (N = 88) |
Northeast | 18% (N = 72) |
Political orientation | |
Very liberal | 11.25% (N = 45) |
Somewhat liberal | 20.25% (N = 81) |
Moderate | 37% (N = 148) |
Somewhat conservative | 18.75% (N = 75) |
Very conservative | 12.75% (N = 51) |
2.2 Design and procedure
The survey respondents were randomly assigned a vignette that described a case and trial involving a physical assault committed by a perpetrator (Jim) against a victim (Ron). The vignette text can be found in Supporting Information S1: Appendix A. Although all respondents read the vignette, one piece of evidence was experimentally manipulated. Respondents were randomly provided one of four pieces of expert evidence during the trial about the age at which Jim experienced physical childhood abuse by his father: prenatally, early childhood, middle/late childhood, or teen adolescence (the independent variable). Physical abuse was chosen for this study as it is the most prominent forms of child abuse, following neglect; neglect is sometimes seen as an act of omission in the care of a child, whereas abuse may be viewed as an act of commission (Gonzalez et al., 2024). Physical abuse and family violence are also the two adverse childhood experiences that are most highly correlated with serious/violent youth offending, and potentially adult offending (Fox et al., 2015).
The text for each condition can also be found in Supporting Information S1: Appendix A. Each condition ended the same way, stating that the abuse only happened during the noted age period and that people like the defendant who were exposed to physical abuse during this time may be aggressive and impulsive as adults. This was important to keep constant so that the age periods were distinct and participants did not have other explicit information to weigh when considering how his abuse could have impacted Jim's adult behavior. A manipulation check during a pilot of 50 respondents was performed and significant differences between the different age categories were observed.
Finally, two additional manipulations were included in the study but are not the focus of the current study. The first manipulation was whether the victim was described as having Down's syndrome, an intellectual disability (Bull, 2020). This manipulation was included to vary perceptions of the victim's “moral patiency,” which is the capacity to be a target of a perceived moral status (Gray & Wegner, 2009). The second manipulation was whether the defendant expressed remorse in the courtroom. Displays of remorse have been linked to sentencing preferences and moral character (Silver & Berryessa, 2023). Models were run controlling for these additional manipulations by including these variables in the models alongside the variable of interest, and the text for these manipulations can be found in Supporting Information S1: Appendix B. 2.3 Measures
After reading the vignette, respondents were asked to provide their support for five goals of Jim's potential criminal sentence (the primary outcome variables). These five goals were rehabilitation, restoration, deterrence, retribution, and incapacitation: rehabilitation aims to reduce an offender's likelihood of future crime commission using treatment and other tools; restoration takes into consideration the needs of the victim and promotes individual responsibility to become a better member of society; deterrence uses the threat of punishment to prevent the offender and others from crim; retribution promotes the idea that the offender needs to be punished because they chose to harm another person; and incapacitation involves physically removing the offender from society to prevent future offending (Miethe & Lu, 2004; Xie et al., 2022).
Adapted from Berryessa (2018), respondents rated the importance of each of the following statements about the goals of Jim's punishment from 1 to 5 (1 being very unimportant and 5 being very important)
Rehabilitation: Helping Jim to turn his life around and stop committing crime.
Restoration: Giving Jim a chance to reconcile with Ron and the community more broadly.
Deterrence: Discouraging Jim or others from committing similar crimes by showing that crimes like Jim's get punished.
Retribution: Making Jim pay for his crime.
Incapacitation: Keeping Jim from committing any more crimes by removing him from society.
Respondents were also asked to rate their opinions on items that tapped into their general levels of trait social essentialism and biological essentialism (see Supporting Information S1: Appendix A for all items). Shortened, amended versions of validated scales were used due to the length of the survey. The social essentialism measure was adapted from the previously validated measure developed by Rangel and Keller (2011), with respondents asked how much they agreed (1 being strongly disagree and 5 being strongly agree) with statements about people in general. A composite, averaged score for participants' trait social essentialism was created and included items A, B, D, and E, with a Cronbach's alpha of 0.72. Statement C was removed because of poor fit with the other items. The biological essentialism measure was adapted from the previously validated measure developed by Keller (2005), with respondents asked how much they agreed (1 being strongly disagree and 5 being strongly agree) with statements about people in general. Statements C and D were used to make a composite averaged score for participants' trait biological essentialism measure, with a Cronbach's alpha of 0.70. Statements A, B, and E were left out due to poor fit.
2.4 Analysis plan
This study used STATA programming software to perform Ordinary Least Squares (OLS) regression to assess the main effects of the age of a defendant's childhood abuse on participants' level of support for each goal of sentencing (rehabilitation, restoration, deterrence, retribution, incapacitation). Then, participants' composite averaged scores for social and biological essentialism measures were included as moderators. An interaction effect model was also used to see if both measures of trait essentialism influenced each independent variable (age of childhood abuse) separately.
In all models, the reference category used for analysis was the prenatal abuse condition. This condition was chosen as the reference category because it is likely that minimal social influences may occur alongside abuse when it occurs prenatally—with the idea that a person's abuse may be viewed or considered to especially shape someone and their essence (“who they are”) if it occurs before they are born (Hiscox et al., 2023). Indeed, with abuse happening after birth, factors other than abuse may be more likely to be viewed to bear on or affect development after birth (Lannert et al., 2014). Thus, using the prenatal category as the reference allows for comparing the effects of abuse pre- and post-birth, including at different ages after birth. Direct comparisons between each and all of the age groups was not the focus or considered in the hypotheses of this study and, therefore, was not reported in this paper. However, results from the rotation of the reference category can be provided upon request.
3 RESULTS
Table 2 shows the descriptive statistics for the primary outcome variables for participants' support for punishment (regardless of the version of the vignette they received) as well as the descriptive statistics for the essentialism items. Tables 3–7 show the primary and moderation effects of the age of childhood abuse on participants' level of support for rehabilitation, restoration, deterrence, retribution, and incapacitation. Column 1 shows the results of the basic OLS regressions for each corresponding punishment goal (outcome variable) and the age of the defendant's childhood abuse (independent variable). Column 2 adds participants' social essentialism score as a general moderator to the basic OLS regression. Column 3 adds the interaction between social essentialism and the ages of abuse. Column 4 adds participants' biological essentialism score as a general moderator to the basic OLS regression. Column 5 adds the interaction between biological essentialism and ages of abuse. All models were also run controlling for the additional manipulations of Down's syndrome and remorse, as well as the demographics of the respondents. No significant differences were observed when including these controls. Demographic controls were not included in the models presented in the tables, but models controlling for participant demographics are available upon request.
TABLE 2. Descriptive statistics of main outcome variables and essentialism items.
Mean | Std. Dev. | Min | Max | |
|---|---|---|---|---|
Outcome variable | ||||
Rehabilitation | 1.56 | 0.902 | 1 | 5 |
Restoration | 2.055 | 1.111 | 1 | 5 |
Deterrence | 1.615 | 0.882 | 1 | 5 |
Retribution | 1.593 | 0.817 | 1 | 5 |
Incapacitation | 2.195 | 1.243 | 1 | 5 |
Essentialism items | ||||
Social essentialism 1 | 2.365 | 1.019 | 1 | 5 |
Social essentialism 2 | 2.245 | 0.947 | 1 | 5 |
Social essentialism 3 | 3.368 | 0.985 | 1 | 5 |
Social essentialism 4 | 2.985 | 1.117 | 1 | 5 |
Social essentialism 5 | 2.627 | 1.11 | 1 | 5 |
Biological essentialism 1 | 2.565 | 1.021 | 1 | 5 |
Biological essentialism 2 | 1.933 | 0.777 | 1 | 5 |
Biological essentialism 3 | 2.935 | 0.989 | 1 | 5 |
Biological essentialism 4 | 2.795 | 1.059 | 1 | 5 |
Biological essentialism 5 | 2.905 | 1.065 | 1 |
TABLE 3. Models for rehabilitation as outcome variable.
(1) | (2) | (3) | (4) | (5) | |
|---|---|---|---|---|---|
Prenatal (reference) | |||||
Abuse 2–5 | 0.104 (0.127) | 0.04 (0.124) | 0.022 (0.125) | 0.106 (0.127) | 0.106 (0.128) |
Abuse 8–11 | −0.258* (0.126) | −0.25* (0.122) | −0.253* (0.122) | −0.256* (0.127) | −0.252* (0.127) |
Abuse 14–17 | −0.131 (0.126) | −0.086 (0.122) | −0.086 (0.123) | −0.128 (0.126) | −0.125 (0.126) |
Down's syndrome | 0.077 (0.089) | 0.095 (0.087) | 0.096 (0.087) | 0.074 (0.09) | 0.074 (0.09) |
Remorse | 0.042 (0.04) | 0.062 (0.038) | 0.061 (0.038) | 0.041 (0.04) | 0.044 (0.04) |
Social essentialism | 0.244*** (0.046) | 0.247** (0.087) | |||
Prenatal × Social | |||||
Abuse 2–5 l × Social | 0.07 (0.116) | ||||
Abuse 8–11 × Social | −0.103 (0.125) | ||||
Abuse 14–17 × Social | −0.006 (0.131) | ||||
Biological essentialism | −0.03 (0.045) | −0.099 (0.082) | |||
Prenatal × Biological | |||||
Abuse 2–5 × Biological | 0.177 (0.129) | ||||
Abuse 8–11 × Biological | 0.055 (0.122) | ||||
Abuse 14–17 × Biological | 0.079 (0.121) | ||||
Constant | 1.211*** (0.248) | 1.279*** (0.24) | 1.286*** (0.241) | 1.224*** (0.249) | 1.211*** (0.25) |
(0.248) | (0.24) | (0.241) | (0.249) | (0.25) | |
Observations | 400 | 400 | 400 | 400 | 400 |
R-squared | 0.05 | 0.115 | 0.12 | 0.051 | 0.056 |
Note
: Beta scores are above the standard errors, which are in parentheses. Model (1) (F (8, 391)= 2.59, p= 0.009) is OLS regression between punishment goal and age of abuse. Model (2) (F (9, 390) = 5.66, p= 0.000) adds Social Essentialism as a moderator. Model (3) (F (12, 387) = 4.41, p = 0.000) adds interaction between Social Essentialism and age of abuse. Model (4) (F (9, 390) = 2.35, p= 0.01) adds Biological Essentialism as moderator. Model (5) (F (12, 387) = 1.92, p= 0.03) adds interaction between Biological Essentialism and age of abuse.
Abbreviation: OLS, Ordinary Least Squares.
***p < 0.001; **p < 0.01; *p < 0.05
TABLE 4. Models for restoration as outcome variable.
(1) | (2) | (3) | (4) | (5) | |
|---|---|---|---|---|---|
Prenatal (reference) | |||||
Abuse 2–5 | −0.102 (0.158) | −0.196 (0.151) | −0.214 (0.152) | −0.101 (0.158) | −0.099 (0.158) |
Abuse 8–11 | −0.435** (0.157) | −0.423** (0.149) | −0.427**(0.15) | −0.433** (0.157) | −0.429** (0.157) |
Abuse 14–17 | −0.122 (0.156) | −0.056 (0.149) | −0.049 (0.15) | −0.121 (0.156) | −0.117 (0.156) |
Down's syndrome | 0.191 (0.111) | 0.217* (0.106) | 0.221* (0.106) | 0.189 (0.111) | 0.18 (0.112) |
Remorse | 0.076 (0.049) | 0.106* (0.047) | 0.104* (0.047) | 0.076 (0.049) | 0.084 (0.049) |
Social essentialism
| 0.359*** (0.056) | 0.351** (0.106) | |||
Prenatal x Social | |||||
Abuse 2–5 × Social
| 0.085 (0.142) | ||||
Abuse 8–11 × Social | −0.122 (0.153) | ||||
Abuse 14–17 × Social | 0.045 (0.16) | ||||
Biological essentialism | −0.014 (0.056) | −0.124 (0.101) | |||
Prenatal × Biological | |||||
Abuse 2–5 × Biological | 0.184 (0.16) | ||||
Abuse 8–11 × Biological | 0.063 (0.151) | ||||
Abuse 14–17 × Biological | 0.224 (0.15) | ||||
Constant | 1.775*** | 1.875*** | 1.887*** | 1.781*** | 1.775*** |
(0.308) | (0.294) | (0.294) | (0.309) | (0.31) | |
Observations | 400 | 400 | 400 | 400 | 400 |
R-squared | 0.038 | 0.131 | 0.135 | 0.038 | 0.045 |
Note: Beta scores are above the standard errors, which are in parentheses. Model (1) (F (8, 391) = 1.93, p= 0.05) is OLS regression between punishment goal and age of abuse. Model (2) (F (9, 390) = 6.51, p= 0.000) adds Social Essentialism as a moderator. Model (3) (F (12, 387) = 5.05, p= 0.000) adds interaction between Social Essentialism and age of abuse. Model (4) (F (9, 390) = 1.72, p= 0.08) adds Biological Essentialism as moderator. Model (5) (F (12, 387) = 1.52, p = 0.11) adds interaction between Biological Essentialism and age of abuse.
Abbreviation: OLS, Ordinary Least Squares.
***p< 0.001; **p< 0.01; *p< 0.05.
TABLE 5. Models for deterrence as outcome variable.
(1) | (2) | (3) | (4) | (5) | |
|---|---|---|---|---|---|
Prenatal (reference) | |||||
Abuse 2–5 | 0.012 (0.127) | 0.008 (0.127) | 0.002 (0.128) | 0.015 (0.127) | 0.009 (0.127) |
Abuse 8–11 | 0.055 (0.126) | 0.055 (0.126) | 0.063 (0.126) | 0.057 (0.126) | 0.058 (0.126) |
Abuse 14–17 | 0.036 (0.125) | 0.039 (0.125) | 0.024 (0.126) | 0.039 (0.125) | 0.038 (0.125) |
Down's syndrome | 0.018 (0.089) | 0.019 (0.089) | 0.007 (0.089) | 0.014 (0.089) | 0.013 (0.09) |
Remorse | −0.015 (0.039) | −0.013 (0.04) | −0.013 (0.039) | −0.015 (0.039) | −0.013 (0.04) |
Social essentialism | 0.015 (0.047) | −0.098 (0.089) | |||
Abuse 2–5 × Social | 0.133 (0.119) | ||||
Abuse 8–11 × Social | 0.29* (0.128) | ||||
Abuse 14–17 × Social | 0.024 (0.134) | ||||
Biological essentialism | −0.031 (0.045) | −0.031 (0.081) | |||
Prenatal × Biological | |||||
Abuse 2–5 × Biological | 0.143 (0.129) | ||||
Abuse 8–11 × Biological | −0.082 (0.121) | ||||
Abuse 14–17 × Biological | −0.03 (0.12) | ||||
Constant | 1.742*** (0.247) | 1.746*** (0.248) | 1.739*** (0.247) | 1.755*** (0.248) | 1.741*** (0.248) |
Observations | 400 | 400 | 400 | 400 | 400 |
R-squared | 0.017 | 0.017 | 0.033 | 0.018 | 0.026 |
Note: Beta scores are above the standard errors, which are in parentheses. Model (1) (F (8, 391) = 0.86), p= 0.55) is OLS regression between punishment goal and age of abuse. Model (2) (F (9, 390) = 0.77, p = 0.64) adds Social Essentialism as a moderator. Model (3) (F (12, 387) = 1.09, p = 0.37) adds interaction between Social Essentialism and age of abuse. Model (4) (F (9, 390) = 0.81, p = 0.61) adds Biological Essentialism as moderator. Model (5) (F (12, 387) = 0.86, p= 0.59) adds interaction between Biological Essentialism and age of abuse.
Abbreviation: OLS, Ordinary Least Squares.
***p< 0.001; **p< 0.01; *p< 0.05.
TABLE 6. Models for retribution as outcome variable.
(1) | (2) | (3) | (4) | (5) | |
|---|---|---|---|---|---|
Prenatal (reference) | |||||
Abuse 2–5 | −0.144 (0.115) | −0.164 (0.115) | −0.159 (0.116) | −0.14 (0.115) | −0.143 (0.115) |
Abuse 8–11 | 0.086 (0.114) | 0.089 (0.114) | 0.089 (0.114) | 0.09 (0.114) | 0.091 (0.114) |
Abuse 14–17 | −0.017 (0.113) | −0.003 (0.113) | 0.013 (0.115) | −0.013 (0.113) | −0.014 (0.113) |
Down's syndrome | 0.054 (0.081) | 0.06 (0.081) | 0.063 (0.081) | 0.049 (0.081) | 0.034 (0.081) |
Remorse | −0.08* (0.036) | −0.073* (0.036) | −0.075* (0.036) | −0.08* (0.036) | −0.072* (0.036) |
Social essentialism | 0.075 (0.043) | 0.04 (0.081) | |||
Prenatal × Social | |||||
Abuse 2–5 × Social | 0.02 (0.109) | ||||
Abuse 8–11 × Social | 0.025 (0.117) | ||||
Abuse 14–17 × Social | 0.118 (0.122) | ||||
Biological essentialism | −0.046 (0.041) | −0.079 (0.074) | |||
Prenatal × Biological | |||||
Abuse 2–5 × Biological | 0.089 (0.116) | ||||
Abuse 8–11 × Biological | −0.085 (0.109) | ||||
Abuse 14–17 × Biological | 0.14 (0.109) | ||||
Constant | 2.112*** (0.225) | 2.133*** (0.224) | 2.14*** (0.225) | 2.131*** (0.225) | 2.129*** (0.225) |
Observations | 400 | 400 | 400 | 400 | 400 |
R-squared | 0.055 | 0.063 | 0.065 | 0.058 | 0.069 |
Note: Beta scores are above the standard errors, which are in parentheses Model (1) (F (8. 391) = 2.86, p= 0.004) is OLS regression between punishment goal and age of abuse. Model (2) (F (9, 390) = 2.90, p= 0.002) adds Social Essentialism as a moderator. Model (3) (F (12, 387) = 2.26, p= 0.009) adds interaction between Social Essentialism and age of abuse. Model (4) (F (9, 390) = 2.69, p= 0.005) adds Biological Essentialism as moderator. Model (5) (F (12, 387) = 2.39, p= 0.006) adds interaction between Biological Essentialism and age of abuse.
Abbreviation: OLS, Ordinary Least Squares.
***p< 0.001; **p< 0.01; *p< 0.05.
TABLE 7. Models for incapacitation as outcome variable.
(1) | (2) | (3) | (4) | (5) | |
|---|---|---|---|---|---|
Prenatal × Social | |||||
Abuse 2–5 × Social | −0.074 (0.169) | ||||
Abuse 8–11 × Social | 0.159 (0.182) | ||||
Abuse 14–17 × Social | 0.023 (0.191) | ||||
Biological essentialism | −0.079 (0.063) | −0.124 (0.115) | |||
Prenatal × Biological | |||||
Abuse 2–5 × Biological | 0.178 (0.181) | ||||
Abuse 8–11 × Biological | 0.004 (0.17) | ||||
Abuse 14–17 × Biological | 0.036 (0.169) | ||||
Constant | 2.584*** (0.349) | 2.582*** (0.35) | 2.573*** (0.35) | 2.616*** (0.349) | 2.602*** (0.351) |
Observations | 400 | 400 | 400 | 400 | 400 |
R-squared | 0.016 | 0.016 | 0.02 | 0.02 | 0.023 |
Note: Beta scores are above the standard errors, which are in parentheses. Model (1) (F (8, 391) = 0.78, p= 0.62) is OLS regression between punishment goal and age of abuse. Model (2) (F (9, 390) = 0.70, p= 0.71) adds Social Essentialism as a moderator. Model (3) (F (12, 387) = 0.67, p= 0.78) adds interaction between Social Essentialism and age of abuse. Model (4) (F (9, 390) = 0.87, p= 0.55) adds Biological Essentialism as moderator. Model (5) (F (12, 387) = 0.75), p= 0.71) adds interaction between Biological Essentialism and age of abuse.
Abbreviation: OLS, Ordinary Least Squares.
***p < 0.001; **p < 0.01; *p < 0.05.
As seen in Table 3, there is significantly less support for rehabilitation as a goal of Jim's sentence when he was described as experiencing childhood abuse between the ages of 8–11 as compared to experiencing it prenatally (B = −0.25, SE = 0.12, t = −2.04, p = 0.042, 95% Confidence Interval (CI) [−0.50, −0.01]). Overall, social essentialism predicted higher levels of support for rehabilitation when it came to abuse in a defendant's background (B = 0.24, SE = 0.46, t = 5.35, p < 0.001, 95% CI [0.15, 0.33]). No other findings were significant when looking at rehabilitation as the outcome variable.
As seen in Table 4, there is significantly less support for restoration as a goal for Jim's sentence when he was described as experiencing childhood abuse between the ages of 8–11 as compared to experiencing it prenatally (B = −0.44, SE = 0.16, t = −2.77, p = 0.006, 95% CI [−0.74, −0.13]). Social essentialism also predicted higher levels of restoration when it came to abuse in a defendant's background (B = 0.36, SE = 0.56, t = 6.45, p < 0.001, 95% CI [0.25, 0.47]). No other findings were significant.
For deterrence, seen in Table 5, the interaction between abuse at ages 8–11 and social essentialism on participants' support for deterrence as a goal of Jim's sentencing was significant. However, because of interactions, a Bonferroni-adjusted significance level of 0.05/3 = 0.0167 was applied and this interaction was no longer significant at p = 0.02. No other findings were significant.
Finally, there were no statistically significant findings for any of the models for retribution or incapacitation (found in Tables 6 and 7, respectively).
4 DISCUSSION AND IMPLICATIONS
This study aimed to explore how the age of a defendant's childhood abuse may affect public support for the goals of sentencing and whether levels of trait essentialism may moderate such levels of support. In addition to adding to an existing body of experimental work on the public's views of and support for considering a defendant's childhood abuse during sentencing (Berryessa, 2021; Najdowski et al., 2009; Platania & Konstantopoulou, 2014; Stevenson et al., 2010), three findings stemmed from the study's results.
First, participants showed less support for the goals of rehabilitation and restoration in the defendant's sentencing when he was described as having experienced childhood abuse between the ages of 8–11, compared to when he experienced abuse prenatally. While this was counter to our hypotheses, the notion that the public may place less importance on the rehabilitation and restoration of defendants abused during this period could be linked to their beliefs that a child substantially develops in terms of emotional development, decision-making, and critical thinking skills during this age period (Cinelli, 2022). Indeed, according to the CDC (2019a), this age period is considered middle childhood, and it is a critical developmental period for a child's physical development due to puberty, learning how to mentally and socially cope with peer pressures and heightened emotional sensitivity, and forming more empathy and understanding (Weinstein, 2022). Thus, it is possible that defendants abused during this period may be viewed as especially unable to be helped and rehabilitated later in life, especially compared to defendants who experienced abuse before birth who may still have opportunities to have their development affected by positive external or social factors (Fogel et al., 2023).
Second, also not aligned with our hypotheses, participants' higher levels of trait social essentialism—across all vignette conditions–predicted increased support for the goals of rehabilitation and restoration in sentencing. However, this may be unsurprising when considering the underlying roots or modes of these two goals of punishment. While social essentialist thinking encompasses beliefs that a person's social background, experiences, environment, and how they are raised are the most important and impactful factors in shaping their future behavior, qualities, and “identity” (Rhodes & Mandalaywala, 2017; Soylu Yalcinkaya et al., 2017), rehabilitation and restoration as goals of punishment also rely on social bonds and relationships to be successful (Tyler, 2006). Other people, including family members and those in the community, are often involved in a defendant's rehabilitation; for example, alternative sentencing options, such as treatment for mental health issues or drug use, often involving sponsors, group and individual therapy with counselors, supportive social networks, and medical professionals to help individuals succeed and recover (SAMHSA, 1997). For restoration, community ties, repairing social bonds with victims, and connecting with family and friends are indispensable in reintegrating a defendant back into their community, enhancing feelings of responsibility, and their commitment to avoid future law-breaking (Tyler, 2006).
Overall, those with higher levels of social essentialist thinking may be more likely to place more importance on or show more support for goals of sentencing that involve engaging a defendant's social bonds, environments, and social processes conducive to positive behavioral changes (Rangel & Keller, 2011). Further, these results also suggest that those with stronger trait social essentialism may lean more toward support for utilitarian forms of punishment, which focus on changing the factors or underlying behaviors driving an individual's criminality (Carlsmith et al., 2002).
Third, there were no significant findings related to participants' levels of trait biological essentialism, regardless of the age at which the defendant had experienced abuse or the potential goal of this punishment. As biological essentialism emphasizes the role of biology, genetics, and other factors pre-birth in behavior formation and trait development (Soylu Yalcinkaya et al., 2017), this was surprising—especially since there is significant research to suggest the effects of childhood abuse on brain development across childhood that can lead to later emotional, physical, and intellectual issues (Fogel et al., 2023; Hiscox et al., 2023). However, these results may suggest that laypeople are potentially more likely to view the effects of childhood abuse, regardless of age, as more of a social rather than biological process; this would complement how abuse and its later effects on criminality are viewed in Social Learning and Social Bond Theories in criminology (Widom, 2017).
There are several limitations to note when considering the results of this study. The original measure of trait biological essentialism included multiple items removed during analysis due to poor fit. As an abbreviated version of the original validated scale (Keller, 2005) was used in this study, future studies may consider using the entire scale to assess trait biological essentialism. Further, this study did not measure other factors that could have affected support for the defendant's sentencing, such as perceptions of Jim's dangerousness or redeemability and participants' prior experiences with or exposure to abuse. Future studies should control for or assess these factors and how they may be related to public support for goals of sentencing in cases in which the defendant has previously experienced childhood abuse.
This research also described only one type of crime (physical assault); results may have differed if the crime was more or less severe. A study involving a more severe crime type, such as murder or its trial in a capital case, could have resulted in more leniency towards a defendant who experienced child abuse—whereas a less severe crime, like physical assault, may not have been considered serious enough to weigh such evidence and result in minimal effects. Concurrently, including another crime type in the vignette in which the abuse that the defendant suffered as a child (physical abuse) does not parallel the crime committed as an adult (physical assault) may have garnered different results. Further, this study also only looked at physical abuse in a defendant's background, as opposed to sexual abuse or other forms of neglect. Views on different types of abuse may differentially impact lay support for different sentencing goals, and future studies should explore how a defendant's history of child sexual abuse at different ages also affects lay support for punishment—as well as essentialism's role in shaping those views.
As there was no experimental condition in which the defendant did not experience any abuse during childhood, future research should also include a condition involving no forms of abuse. Additionally, it is unknown to what extent participants and people in the general public have knowledge or understanding about the stages of development during different periods of childhood; including a measure to determine the knowledge that participants have about childhood development would be helpful for work on these issues moving forward.
Another limitation, which was outside of the scope of this study, was that participants beliefs or thoughts on the cause or reason for the abuse were not assessed. This was not described in the vignette, so it is possible that participants viewed the defendant who had experienced prenatal abuse as having less fault in or “cause” in experiencing the abuse compared to those who experienced abuse at other ages. Different assumptions on the causes of abuse at various ages may have influenced the overall results and this should be assessed in future work. Relatedly, future studies may want to incorporate participant beliefs about child development at different age ranges, such as their assessments on developmental processes and abilities of children in each age group. Additionally, we chose to measure participants' levels of trait essentialism after the vignettes; as the information included in the vignettes may have influenced participant levels of essentialism, particularly social essentialism, future studies should consider the order in which measures are presented.
As one final limitation, experimental studies using online surveys and samples are inherently limited regarding ecological validity, which refers to how generalizable the procedures and findings of a study are to the real world (Neuliep, 2017). When people make decisions in real life, they are not making them on a survey or using a scale—and other factors may influence their decisions, as opposed to sitting at a computer with selected information. The individuals in the sample for this research also self-selected and chose to be part of an online survey pool. The use of vignettes also adds to the issue of ecological validity, as it does not necessarily reflect how evidence or other court proceedings would be presented or viewed in an actual court setting. Therefore, future research should replicate this study with other vignettes, constructs, and variables, as well as include more realistic or ecological designs such as a mock trial with video trial recordings and evidence presentation.
In conclusion, this study provides valuable insights for future work and policy. For future empirical work, this study can be used as a starting point for future experimental work on these issues and contribute to the improvement of methods and materials to assess these and similar research questions related to assessing the public's views of sentencing practices. For policy, this study begins to suggest how child abuse may be factored into sentencing practices moving forward. As the U.S. Sentencing Commission has expressed that sentencing practices should, to some degree, align with the public's views on what is appropriate sentencing (Berryessa, 2021; Maxfield et al., 1996; U.S. Sentencing Commission, 1995), the results of this study support prior work to suggest that members of the public consider a defendant's history or background of childhood abuse as important to the consideration of their sentence—mainly related to sentences that aim to rehabilitate a defendant and restore their relationships with victims and the community (Berryessa, 2021). As reflecting the public's views in sentencing practices is crucial in order to better ensure that capital and non-capital sentencing practices are viewed as fair and legitimate (Maxfield et al., 1996), these data suggest the public's interest in integrating child abuse evidence into sentencing considerations, particularly as it relates to weighing mitigation and alternative goals of punishment aside from retribution and incapacitation (Roberts & de Keijser, 2014).
As members of the public, like the participants of this study, may also serve as jurors to determine verdicts and sometimes punishment outcomes in capital and some non-capital (Chalabi, 2015), it is also important to understand the potential factors, such as age of experiencing abuse, that may predict and affect support for sentencing outcomes and practices in cases involving defendants with histories of child abuse. Indeed, by understanding how lay individuals determine decisions of responsibility and sentencing in cases involving defendants with histories of abuse, the system can better attune its practices to foster more positive views of the legal system as a whole and further more procedurally just sentencing practices (Sunshine & Tyler, 2003; Tyler & Lind, 2001).
These results also indicate that having higher levels of trait essentialism beliefs, particularly social essentialism, can impact lay support for sentencing practices in cases involving defendants with histories of child abuse. While gauging the public's views and support for different goals of sentencing are important to ensuring that sentencing practices are viewed as fair and legitimate (Maxfield et al., 1996), this study also adds to existing work to suggest that lay views on sentencing can be commonly affected by implicit and unconscious processes of which they themselves may be largely unaware (Fousiani & van Prooijen, 2019). While essentialism and other unconscious beliefs are exhibited to some degree by all humans and cross-culturally (Bastian & Haslam, 2006; Gowaty, 2018), it is important to be aware of their impacts when contemplating how public views can and should be integrated or considered in future sentencing reform and policy, including in cases involving defendants who have experienced childhood abuse.
In this vein, another more practical policy implication may involve the education of the public and other programs that could reduce the potential negative effects that essentialist attitudes and thinking could have on different defendant outcomes and judgments. Providing forms of lay education on essentialist beliefs and other unconscious biases that play a role in judgments and decision-making can help to mitigate their expression and effects (Daftary-Kapur et al., 2010; Su, 2020); this could be applied to juror and other lay legal education practices. The public may also be educated more generally about essentialism through media and other forms of information about the criminal justice system, as directly presenting people with their own biases and beliefs can reduce their negative consequences (Banaji & Greenwald, 2016). Further, specifically regarding cases involving a defendant with abuse in their childhoods, lawyers may also want look to the results of this study and future studies to inform their voir dire processes, such as asking questions that tease out essentialist beliefs when assessing jurors for different cases.