The Good,the Bad,and the Incomprehensible

Abstract

Public opinion has played a critical role in the development of sex crime laws. However, little scholarly work has focused directly on the origins of negative attitudes toward sex offenders. We address this research gap by developing and testing a theoretical account of such views. Drawing on recent national survey data, we examine the extent to which typifications about sexual victims and offenders—believing sex crime typically affects children and female victims and is committed by strangers—explain beliefs about the reformability of sex offenders, harm inflicted on victims, and the causes of offending. Results indicate that judging children to be typical targets of sex crimes is a key determinant of public views. We discuss the implications of our findings.

Keywords

crime perceptions causal attributions sex offending victimization

Introduction

Since the appearance of “get tough” sex crime laws in the mid-1990s, states and the federal government have continued to implement an array of new punitive reforms. As a result, the post-incarceration sanctions that apply to sex offenders in contemporary society are numerous. Registration, community notification, residence restrictions, chemical castration, and civil commitment are examples of laws enacted nationally over the last two decades. The perception that sex offenders represent the “worst of the worst” offenders is clearly implicit in the content of these laws given that (a) they apply exclusively to sex offenders (Ackerman, Sacks, & Greenberg, 2012; Jenkins, 1998; Lytle, 2013), and (b) assume that sex offenders are compelled to reoffend, and so are in need of intense supervision and monitoring following release (Meloy, Curtis, & Boatwright, 2013).

Remarkably, the proliferation of sex crime laws in the United States has not followed an increase in actual sex offending. Indeed, sex offenses involving both child and adult victims have substantially decreased in the United States. In analyzing trends for child sexual victimization, Finkelhor and Jones (2012) have contended that the decline is “about as well established as crime trends can be in contemporary social science” (p. 3). In a similar direction, the most recent data available indicate that reports of forcible rape—an indicator of sex crime involving older and adult victims—remain at record low levels (Planty, Langton, Krebs, Berzofsky, & Smiley-McDonald, 2013). If not rising sex crime rates, what factors are associated with the pronounced change in sex offender management? Some scholars have identified public opinion, specifically, increased concern about sexual victimization as a potential catalyst in the emergence of punitive sex crime reforms (Fox, 2013; Pickett, Mancini, & Mears, 2013).

For example, there are widely held societal impressions, identified as early as 1950 by Edwin Sutherland, surrounding the nature of sex crime. Perhaps the most prominent of these centers on the reformability of sex offenders. In Sutherland’s (1950) seminal article criticizing sexual psychopath laws (i.e., an early form of civil commitment), he cited a statistic proffered by J. Edgar Hoover that “the most rapidly increasing type of crime is that perpetrated by degenerate sex offenders” (p. 543). Elsewhere, Sutherland explained that the typical perception underpinning sex crime laws is the view that “sexual psychopaths continue to commit serious sex crimes throughout life because they have no control over their sexual impulses” (p. 544). This incorrigibility argument has stood largely unchallenged for over 60 years. To illustrate, in a recent national poll nearly three quarters of the public expressed that sex offenders could never be rehabilitated (Mancini, 2014). Separately, Sutherland alluded to another proposition—the belief that sexual victimization imparts greater relative harm to potential victims than other crimes. In contemporary times, sexual victimization has been judged to inflict “psychological trauma [that] has no end point” (Schafran, 1992, p. 446). In contrast to other offenses, it is a crime that simultaneously “robs [victims’] innocence” (Harbeck, 2009, p. ix), “maim[s] [their] soul[s]” (Schafran, 1992, p. 439), “kills their spirit” (Zilney & Zilney, 2009, p. 96), and “envelops [victims] in[to] toxic shame” (Harbeck, 2009, p. ix). Accordingly, sexual victimization, in comparison with other violent offenses, is perceived as an offense that inflicts everlasting emotional turmoil on victims. Third and finally, Sutherland recognized that there are strong views related to the causes of sex offending. He pointed to an array of typical descriptors of sex offenders—“degenerates,” “sex fiends,” “creatures,” and “sexual psychopaths” (p. 543)—all of which, per Sutherland, highlight that sex offending is perceived as pathology or a dispositional “mental malady”—in turn, the only solution would be to “segregate such persons preferably before, but at least after their sex crimes” (p. 544). Here again, public opinion has not wavered, that is, most Americans typically attribute individual deficits—selfishness and moral depravity (e.g., Spencer, 2009)—as opposed to environmental or situational factors to sex offending.

In a word, there are certain widely held impressions about sex crime—that sex offenders are unreformable, inflict enduring suffering on their victims, and commit their offenses because of dispositional deficiencies, such as personal immorality and low self-control—that collectively foster public hostility toward this group of criminals. Indeed, recent research shows that these impressions constitute the three strongest predictors of punitive attitudes toward sex offenders (Pickett et al., 2013). Even so, left unaddressed by extant work is research investigating the origins of these views. To illustrate, Payne, Tewksbury, and Mustaine (2010) observed that although “it is clear that sex offenders are stigmatized and this stigma has hindered appropriate responses to sex offenders,” at the same time, it is “not clear how this stigma arises” (p. 587). The theory advanced in the current study is that these stigmatizing beliefs about sex offenders are consequences of the public’s intense intuitive revulsion toward (see Haidt, 2001), and general inability to understand, sexual offending, rooted in two specific typifications—or collective images—about sex offenders and victims.

First, sex crime is perceived as disproportionately affecting a victim base not targeted by any other type of offender—vulnerable children and women (Meloy et al., 2013; Sample & Kadleck, 2008). It is likely that offenses against this victim base trigger intense moral emotions and, in the public’s view, cannot extend from “normal” motives, such as economic gain or interpersonal anger, and thus those who commit them cannot be normal individuals. Second, there is the widely endorsed perception that victims are at substantial risk of being assaulted by stranger perpetrators, a view reflecting the myth of “stranger danger” (Craun & Theriot, 2009) and contributing to the perception of sexual criminals as persons who are outside of and stand apart from the normal community (see Spencer, 2009).

Collectively, we argue that these contemporary images of sex crime give rise to and reinforce the negative beliefs about sexual offenders that underlie public hostility toward sex criminals. Specifically, we hypothesize that these perceptions, by triggering disgust, rendering sex crime incomprehensible, and undermining the ability of the public to conceive of sexual offenders as members of the community, cultivate impressions about offender incorrigibility, sex victim suffering, and the immoral dispositions of sex criminals. Using data from a recent national survey, we provide the first test of this prediction in the current study.

Societal Impressions of Sex Crime Victims and Sex Offenders

The 1980s ushered in a shift in public discourse about crime. Specifically, the victim rights movement resulted in laws and policies designed to protect victims of crime and involve them in criminal proceedings (Garland, 2001). The end result per Garland and Sparks (2000) is that

citizens [have become] crime-conscious, attuned to the crime problem, and many exhibit high levels of fear and anxiety . . . caught up in institutions and daily practices that require them to take on the identity of (actual or potential) crime victims, and to think, feel, and act accordingly. (p. 200)

This emphasis is clearly exemplified when the focus is on sex crime victims. Specifically, in contemporary society, certain groups—namely children and women—have been portrayed as particularly vulnerable to sexual violence from dangerous predators (Kitzinger, 2004). Levenson, Brannon, Fortney, and Baker (2007) reported that over 50% of the public felt registries should publicize the age of offenders’ prior victims—in turn, suggesting that Americans prioritize such information when judging the disposition and dangerousness of sex offenders. Relatedly, it is largely recognized by the public that women are disproportionately victims of forcible rape and sex offenses (Best, 1999). At the same time, “get tough” sex crime laws appear to rest on a “vulnerable protection” rationale. For instance, residence restrictions typically prohibit offenders from living near areas frequented by children, such as schools, daycare centers, and playgrounds (Anderson, Sample, & Cain, 2013). Sample and Kadleck (2008) in their qualitative analysis examining legislators’ views about sex offenders remarked, “to no surprise, women and children were most frequently mentioned [by respondents as victims of sex crime]” (p. 52). Collectively, such observations create the impression that sex crime laws exist to protect vulnerable populations.

Strikingly, this emphasis is not portrayed in accounts of other violent and serious crimes (Greer, 2007). Put differently, sex offending is perceived as affecting a unique victim base—one that is typically not the target for other offenses (e.g., robbery). The pithy adage, “Many a man that needed killin’, no child that needed a molestin’” (Nhan, Polzer, & Ferguson, 2012, p. 829), succinctly demonstrates the difficulty that the public has in understanding sexual offending—a crime that disproportionately involves vulnerable victims—vis-à-vis other offenses.

We argue that the incomprehensibility of crimes against vulnerable persons triggers intuitive moral emotions and, consequently, fosters negative cognitive judgments about sexual offenders along a range of dimensions. For example, Haidt (2001) hypothesized that moral judgments about social phenomenon are typically the result of “quick, automatic evaluations (intuitions)” rather than careful, rational reflection and deliberation (p. 814). As a result, such negative snap judgments have the tendency to induce assessments of social harm arising from a particular action or behavior. Under this logic, endorsing typifications about populations portrayed as vulnerable to sexual victimization may trigger a negative “intuitive repulsion” (i.e., an instant and intense feeling, see Haidt, 2001; Haidt, Koller, & Dias, 1993) leading to the endorsement of faulty cognitive judgments about sex crime, particularly concerning the relative harm of sexual victimization. In addition, psychological research indicates that because individuals have an intuitive need for developing coherent explanatory narratives, they tend to exaggerate “the consistency of evaluations: good people do only good things and bad people are all bad” (Kahneman, 2011, pp. 199-200). The perception that sex offenders intentionally target vulnerable populations is thus likely to undermine the ability of individuals to attribute to sex offenders any favorable traits or the potential for prosocial behavior—such as the potential for reform, or the possibility that their harmful behavior is environmentally motivated (i.e., due to situational factors rather than dispositional traits). Using this logic as a springboard for the current study, we hypothesize the following:

Hypothesis 1: Individuals who perceive that a larger percentage of sex crime victims are children will be more likely to believe that sex offenders are unreformable, inflict greater relative harm on victims, and offend because of dispositional factors.

Hypothesis 2: Those who perceive that a larger percentage of sex crime victims are female will be more likely to believe that sex offenders are unreformable, to judge that sexual victimization inflicts greater relative harm on victims, and to feel that sex offenders offend due to dispositional factors.

Separately, the “stranger danger” myth has been featured prominently in accounts of sex offending. A number of public opinion polls indicate that a non-trivial number of Americans endorse the view that victims and offenders are typically unfamiliar with one another (Fuselier, Durham, & Wurtele, 2002; Levenson et al., 2007). In a similar direction, Zgoba (2004) reported that although stranger perpetrated offenses comprise a very small proportion of total sex crimes (2% of all crimes against juveniles per her estimates), a guiding belief behind sex crime laws is that most victims are at risk of victimization by stranger offenders. Put differently, endorsed by the public and implicit in sex crime legislation is the stranger danger myth. Here again, we argue that this perception, by strengthening the conception of sex offenders as monstrous outsiders (Spencer, 2009), cultivates broader negative views about sex crime by intensifying the public’s negative emotional reaction to, and thus inability to comprehend, sex crime. Specifically, we present our final hypothesis below:

Hypothesis 3: Those who perceive that a larger percentage of sex crimes are committed by strangers will be more likely to believe that sex offenders are unreformable, to judge that sexual victimization inflicts greater relative harm on victims, and to feel that sex offending is driven by dispositional factors.

Data and Methods

Sample and Data Collection

We rely on data collected from an online survey administered through SurveyMonkey and conducted in the summer of 2012. Our sampling frame for the current study included an opt-in panel (SurveyMonkey’s Audience panel) of more than 400,000 individuals,¹ from which a sample of 2,898 people were randomly selected to receive an e-mail invitation to participate. The opt-in panel consists of individuals who volunteer to complete surveys in exchange for various incentives; respondents can receive automatic donations in their name to charitable organizations and opportunities to win $100 in weekly drawings.

Because they are demographically diverse and produce high-quality data (Bhutta, 2012; Buhrmester, Kwang, & Gosling, 2011; Gosling, Vazire, Srivastava, & John, 2004; Sanders, Clarke, Stewart, & Whiteley, 2007), opt-in online samples, including those drawn from the Audience panel, are increasingly being used in published research to examine social issues and public attitudes, including studies of crime and justice (see Blodorn & O’Brien, 2013; Griswold & Wright, 2004; Pickett & Baker, 2014). SurveyMonkey permits researchers to specify targeting criteria for the random selection of respondents. For our study, only adults (i.e., those 18 or older) and U.S. residents were recruited to participate. In addition, given that African Americans tend to be underrepresented in web-based surveys (Bethlehem & Biffignandi, 2012), an attempt was made to oversample this population. No information regarding the specific content of the survey was provided to potential participants in the initial invitation e-mail.² A total of 612 Americans responded to the solicitation. Approximately 88% (n = 537) of those who began the survey completed the questionnaire.

The final sample included respondents from every state, but South Dakota. Respondents were not concentrated in any particular region, but rather exhibited diversity in regional residence: 18% resided in the Northeast, 22% in the Midwest, 34% in the South, and 26% in the West. In terms of demographics, the sample was generally similar to the U.S. population of adults with a few exceptions. Specifically, the sample had the following characteristics (percentages for the sample and national estimates presented parenthetically): women (51% of the sample vs. 51% nationally), White (75% of the sample vs. 75% nationally), Latino/Hispanic (5% of the sample vs. 14% nationally), African American (16% of sample vs. 12% nationally), 65 years or older (15% of the sample vs. 17% nationally), college educated (57% of the sample vs. 25% nationally), and annual income of $100,000 or more (27% of the sample vs. 21% nationally; see U.S. Census Bureau, 2010, 2012, for national estimates). Thus, with the exception of differences in racial/ethnic composition, education, and income, the sample approximated the demographics of the larger U.S. population. Notably, such discrepancies are typical in web-based survey research (see Chang & Krosnick, 2009). Furthermore, we do not anticipate these differences to affect results given that (a) prior research has generally found that respondents’ race/ethnicity, education, and income are either unrelated or weakly linked to views about sex offenders (Button, Tewksbury, Mustaine, & Payne, 2013; Mancini, 2014; Payne et al., 2010), and (b) we control for these factors in the models (see Winship & Radbill, 1994).

Drawing on web-based survey data offers several advantages. First, online surveys provide greater flexibility than other methods (e.g., telephone surveys)—permitting respondents to complete the survey at their convenience (Rhodes, Bowie, & Hergenrather, 2003). Second, web-based surveys can be designed to be dynamic and user friendly by including pop-up instructions, drop-down boxes, check boxes, and skip patterns (Bethlehem & Biffignandi, 2012). In turn, these design features may increase respondents’ motivation to complete the survey (Schmidt, 1997; Zhang, 1999). Separately, because online surveys are self-administered, social desirability bias is reduced (Tourangeau & Yan, 2007), and thus, more valid responses to sensitive questions are elicited. Moreover, coding errors are minimized as respondents directly record their answers to survey questions (Zhang, 1999).

To be sure, like any polling method (e.g., telephone, mail surveys), there are limitations to web-based surveys. One involves coverage, that is, respondents must have Internet access to participate in online surveys. Another limitation, which is partially a consequence of the first is that the typical online survey participants tend to have higher incomes and greater educational attainment than the general public (Gosling et al., 2004). Even so, the use of an online sample is appropriate for our purposes because our focus is on providing an initial test of theoretical hypotheses rather than on estimating the prevalence of attitudes or behaviors in the general population.

Measures

Dependent variables

Reformability

The perception that sex offenders are driven to reoffend reflects one of the most enduring, yet, strikingly, empirically unsupported societal impressions of sex crime (Mancini, 2014; Sutherland, 1950). Accordingly, our first outcome variable measures the extent to which the public judges sex offenders to be unreformable. In line with prior research (e.g., Payne et al., 2010), we rely on the following measure: “In general, do you think sex offenders³ can be successfully rehabilitated?” Response options were 0 = yes and 1 = no.

Relative harm

Separately, another prominent perception held about sex crime is that sexual victimization inflicts a substantially greater extent of everlasting harm on victims relative to other violent crimes (Harbeck, 2009; Schafran, 1992; Zilney & Zilney, 2009). Even so, virtually no research has evaluated the extent to which the public endorses this view, or by extension, the origins of such judgments. Here, we tap into this perception by creating a relative harm index that includes seven items (α =.88) measuring beliefs about the relative harm caused by sexual victimization.⁴

Dispositional cause of sex offending

The belief that certain dispositional factors of offenders—selfishness and immorality—drive sex offending is a view that has withstood the test of time (compare Spencer, 2009; Sutherland, 1950). However, the origins of this belief are not clear. Accordingly, we include this perception as our last dependent variable by creating an index derived by averaging responses to three items: (a) “most sex offenders commit sex crimes because they have bad moral character,” (b) “most sex offenders commit sex crimes because they have been exposed to pornography in the past,” and (c) “most sex offenders commit sex crimes because they are just selfish people.” The three items loaded on one factor (eigenvalue = 1.72; loadings ranged from 0.68 to 0.80; α = .62).⁵ Responses were coded as 1 = strongly disagree to 5 = strongly agree. Thus, higher values indicate greater agreement that dispositional factors (e.g., offender immorality) lead to sex offending.

Independent variables

Sex crime typifications

Certain perceptions—exclusive to sex offenders—are widely endorsed by a large swath of the public. As discussed earlier, two specific views appear prominent. First is the impression that vulnerable populations—children and women—are disproportionately targeted by sex offenders. To measure typifications about victims and offenders, prior studies typically ask respondents to estimate the percentages (0%-100%) of these populations who possess the characteristic of interest (e.g., Levenson et al., 2007; Pickett & Chiricos, 2012). Our survey included two such items that tap the relevant typifications about sex crime victims. First, respondents were asked, “When you think about the victims of sex crimes, approximately what percent would you say are young children, what percent are teenagers, and what percent are adults?” Then, the survey queried, “When you think about the victims of sex crimes, approximately what percent would you say are female and what percent are male?” The two variables are equal to the percentages reported for young children and females, respectively.

Another consistent impression is that sex offenders are typically stranger perpetrators. The survey included the following question: “When you think about the victims of sex crimes, approximately what percent would you say are related to the offender, what percent know the offender but are not related, and what percent don’t know the offender?” The resulting variable is equal to the percentage reported for the “stranger” choice (i.e., “don’t know the offender”).

Controls

In line with similar studies (Anderson et al., 2013; Payne et al., 2010), several control variables were included in the analyses. Sex was self-reported (0 = male, 1 = female). Respondents also provided their parental status (0 = no children under age 18, 1 = child under age 18). Marital status was measured as 0 = not currently married and 1 = married. Race was coded dichotomously (0 = non-White, 1 = White). Age was recorded as follows: 1 = 18 to 24 years of age, 2 = 25 to 34 years of age, 3 = 35 to 44 years of age, 4 = 45 to 54 years of age, 5 = 55 to 64 years of age, and 6 = 65 years of age or older. Education was measured in categories (1 = high-school diploma or less, 2 = some college, 3 = bachelor’s degree, 4 = master’s degree, law degree, or similar graduate degree, and 5 = PhD, MD, or other advanced graduate degree). Income included the total household income reported for 2011, where 1 = up to $14,999, 2 = $15,000 to $34,999, 3 = $35,000 to $49,999, 4 = $50,000 to $74,999, 5 = $75,000 to $99,999, and 6 = $100,000 or more.⁶ To tap into conservatism, the survey asked: “How would you describe yourself politically?” (higher responses indicate greater conservatism): 1 = very liberal, 2 = liberal, 3 = middle of the road, 4 = conservative, and 5 = very conservative. Religiosity was gauged by asking: “How important would you say religion is in your life?” (1 = very unimportant, 2 = unimportant, 3 = somewhat unimportant, 4 = somewhat important, 5 = important, and 6 = very important).

Two specific controls related to sex crime perceptions were also incorporated in analyses. The logic here is that such perceptions may contribute to negative assessments of sex offenders, and so their effects should be accounted for in the models. First, perceived sex crime trend was created from responses to this question: “In your best judgment, how has the number of sex crimes committed annually in the United States changed over the past 5 years?” Original response categories were 1 = decreased greatly, 2 = decreased some, 3 = stayed the same, 4 = increased some, and 5 = increased greatly. As no respondent volunteered that sex crimes had “decreased greatly,” the variable was coded as follows: 1 = decreased, 2 = stayed the same, 3 = increased some, and 4 = increased greatly. Second, the survey included a question that gauged respondents’ perceived lack of control in avoiding sexual victimization: “How much control do you feel you have over whether or not someone in your family becomes the victim of a sex crime in the next 5 years?” Respondents could choose 1 = a great deal, 2 = a good amount, 3 = some, 4 = very little, and 5 = none at all.

We were also able to determine whether anyone in the respondents’ immediate family had ever been sexually victimized. We coded this variable—sex crime victim—as 0 = no sexual victimization and 1 = sex crime victim. In addition, respondents were asked to evaluate the level of community disorder in their neighborhoods. We relied on an eight-item question similar to those used in prior studies (e.g., Sampson & Raudenbush, 2004; Skogan, 1992) to gauge perceptions of disorder.⁷ In each instance, respondents could report, 1 = not a problem, 2 = a small problem, 3 = a problem, 4 = a big problem, and 5 = a very big problem. The responses to these eight questions were averaged to create an index. The Cronbach’s alpha value (α = .89) indicates high internal consistency among the measures.

Not least, given prior hypotheses linking local area characteristics with public opinion about sex offenders (Button et al., 2013; Payne et al., 2010), we include four contextual controls. The percentage of individuals younger than 18 residing in the respondents’ zip codes (measured using U.S. Census data) was included given arguments regarding vulnerable populations. Analyses also account for the number of registered sex offenders (RSOs) living in the respondents’ zip codes (measured using data from the Dru Sjodin National Sex Offender Public Website). Because the original distribution for this variable was skewed, we use its square root. The percentage of Republicans living in the respondents’ zip codes was also included given arguments regarding conservatism and public opinion (Unnever, Cullen, & Fisher, 2007). Here, percent Republican is equal to the percentage of the county that voted for John McCain in the 2008 presidential election. Finally, in line with hypotheses regarding region and public views (Borg, 1997), analyses control for the geographical region in which the respondent resides (0 = non-South, 1 = South). Table 1 provides descriptive statistics for all of the variables used in the models. Below, we detail our multivariate analysis.

Table 1.

Descriptive Statistics.

Variables	M	SD
Dependent variables
Offender reform	0.57	0.50
Relative harm to victims	3.43	0.59
Assessment of causes of sex offending	3.20	0.88
Independent variables
Perceived percent young	38.49	17.97
Perceived percent female	73.43	15.86
Perceived percent stranger	23.62	16.67
Control variables
Individual characteristics
Female	0.51	0.50
Parent	0.28	0.45
Married	0.58	0.49
White	0.75	0.43
Age	3.89	1.51
Education	2.84	1.07
Income	4.09	1.58
Conservatism	2.83	1.03
Religious	3.77	1.85
Sex crime increase	2.78	0.90
Perceived control	3.08	1.07
Sex crime victim	0.26	0.44
Community disorder	1.72	0.75
Local area characteristics
Percent younger than 18	23.55	3.01
Sqrt number of RSOs	5.36	3.26
Percent Republican	42.01	14.52
South	0.34	0.47

Note. Sqrt = square root; RSO = registered sex offenders.

Analytic plan

Multivariate analysis proceeded in three steps. First, given the dichotomous nature of the first outcome variable—offender reformability—we estimated logistic regression models (see Table 2) that first accounted for control variables (Model 1) and then included our predictors of interest (Model 2). This strategy permitted us to evaluate the extent to which the theoretical variables improved model fit net of the controls. Second, we relied on ordinary least squares (OLS) regression for the second set of models (see Table 3) which examined predictors of perceptions of relative harm. Here again, the first model included the control variables and the second accounted for the control variables and the main predictors. The third set of models utilized OLS regression and focused on the last outcome measure—perceived causes of sex offending. In line with the previous runs, we estimated two models—one designed to account for demographic and local area characteristics and the other included the predictors of interest plus controls. Cases with missing values for all variables, except income, were dropped from analysis using listwise deletion. Because multilevel modeling was not possible, robust standard errors, cluster by county, were used to correct for potential autocorrelation in the residuals.⁸

Table 2.

Logistic Regression Analyses of the Effects of Societal Images on Public Perceptions of Offender Reform.

	Model 1			Model 2
Variables	b	RSE	St. Coef.	b	RSE	St. Coef.
Societal images
Perceived percent young	—	—	—	.014*	(.006)	.126
Perceived percent female	—	—	—	−.008	(.007)	−.059
Perceived percent stranger	—	—	—	.006	(.007)	.050
Control variables
Female	.544**	(.210)	.135	.578**	(.218)	.142
Parent	.338	(.258)	.076	.295	(.261)	.066
Married	−.391	(.244)	−.096	−.357	(.249)	−.086
White	.503^†	(.275)	.108	.582*	(.282)	.123
Age	.164*	(.076)	.123	.142^†	(.076)	.104
Education	−.010	(.098)	−.006	.037	(.099)	.019
Income	−.021	(.078)	−.017	−.015	(.078)	−.012
Conservatism	.209^†	(.121)	.106	.222^†	(.124)	.111
Religiosity	.031	(.066)	.029	.011	(.068)	.010
Sex crime increase	.554***	(.113)	.249	.548***	(.114)	.243
Perceived control	.139	(.090)	.074	.142	(.093)	.075
Sex crime victim	.372	(.261)	.082	.413	(.267)	.090
Community disorder	.195	(.146)	.074	.175	(.143)	.065
Percent younger than 18	.025	(.042)	.038	.023	(.041)	.034
Sqrt number of RSOs	−.033	(.031)	−.053	−.031	(.031)	−.050
Percent Republican	−.009	(.009)	−.065	−.009	(.009)	−.067
South	−.140	(.232)	−.033	−.199	(.234)	−.047
Nagelkerke’s R²		.182			.202

Note. b = unstandardized regression coefficient; RSE = robust standard error; St. Coef. = standardized regression coefficient; Sqrt = square root; RSO = registered sex offender.

†

p ≤ .10. *p ≤.05. **p ≤ .01. ***p ≤.001 (two-tailed significance test).

Table 3.

OLS Regression Analyses of the Effects of Societal Images on Public Perceptions of Relative Harm.

	Model 1			Model 2
Variables	b	RSE	St. Coef.	b	RSE	St. Coef.
Societal images
Perceived percent young	—	—	—	.004**	(.002)	.133
Perceived percent female	—	—	—	.001	(.002)	.022
Perceived percent stranger	—	—	—	−.002	(.002)	−.044
Control variables
Female	.198***	(.056)	.167	.196***	(.056)	.166
Parent	−.008	(.062)	−.006	−.024	(.062)	−.019
Married	.038	(.064)	.031	.038	(.063)	.031
White	−.000	(.073)	−.000	.022	(.072)	.016
Age	−.039^†	(.022)	−.099	−.046*	(.022)	−.116
Education	.011	(.029)	.019	.021	(.030)	.038
Income	−.006	(.020)	−.015	−.004	(.020)	−.011
Conservatism	.045	(.028)	.078	.045	(.028)	.078
Religiosity	−.009	(.015)	−.029	−.009	(.015)	−.029
Sex crime increase	.155***	(.031)	.237	.146***	(.031)	.223
Perceived control	.041^†	(.022)	.073	.032	(.022)	.057
Sex crime victim	.014	(.063)	.010	.016	(.063)	.012
Community disorder	−.093	(.059)	−.119	−.096	(.060)	−.123
Percent younger than 18	.006	(.012)	.031	.006	(.011)	.030
Sqrt number of RSOs	−.012	(.008)	−.066	−.012	(.008)	−.066
Percent Republican	−.003	(.002)	−.081	−.004^†	(.002)	−.089
South	−.027	(.054)	−.022	−.031	(.054)	−.025
R ²		.127			.146
Adjusted R²		.096			.110

Note. OLS = ordinary least squares; b = unstandardized regression coefficient; RSE = robust standard error; St. Coef. = standardized regression coefficient; Sqrt = square root; RSO = registered sex offender.

†

p <.10. *p < .05. **p < .01. ***p <.001 (two-tailed significance test).

Results

Starting with our first area of examination, do certain typifications of victims and offenders drive negative judgments about sex offenders’ ability to be reformed? Results from Table 2 suggest “yes” and “no.” Specifically, the belief that a larger percentage of children are most typically victims of sex crime was associated with negative assessments of offender reform (β = .126). Notice here that this effect is relatively large in magnitude. As inspection of Table 2 indicates, it is the third strongest predictor of reformability perceptions. However, the typification that women are predominantly victims of sex crime was not linked with reformability views. Separately, the perception that strangers perpetrate most sex crimes constitutes another prominent belief endorsed by a large swath of the public and one that drives public policy. Even so, it is not a view that significantly predicts perceptions of offender reform. Thus, findings from Table 2 support the study’s first hypothesis, but lend no support to the second or third hypothesis. The addition of the theoretical predictors improved model fit (Nagelkerke’s R² increased from .182 to .202). Other significant findings were evident. Females (β = .142), Whites (β = .123), older individuals (p ≤ .10; β = .104), those with greater conservatism (p ≤ .10; β = .111), and those who perceive an increase in sex offenses (β = .243) were more likely to feel sex offenders could not be reformed. Of these significant results, endorsing the child typification view, being female, and perceiving an increase in sex crime exert the strongest effects in the model.

Do these predictors influence views about the relative harm of sex victimization? Here again, mixed support is found in Table 3. Only the child victim typification is linked with a stronger belief that sex crime imparts additional harm on victims compared with other violent crimes. Again, this effect is not trivial in magnitude—child victim typification is among the strongest predictors in the model (β = .133). The other two typifications—female and stranger beliefs—do not appear to shape views about relative harm. Support then is found for our first hypothesis, but not for the other two. The inclusion of these predictors improves model fit (adjusted R² increased from .096 to .110). In addition, four controls bear mention. Specifically, sex (β = .166), age (β = −.116), believing that sex crime has increased (β = .223), and percent Republican (p < .10; β = −.089) were also significant or marginally significant predictors. In particular, the child victim belief, sex, and perceiving an increase in sex crime had the strongest effects on relative harm views.

The final set of models (shown in Table 4) analyzed factors related to perceived causes of sex offending. In contrast to the other two outcomes, here only the female typification predicted views about the causes of offending, and only at the .10 significance level (β = .085). The other two typifications exerted no significant effect on perceptions of sex crime causes. This finding lends some support to the study’s second hypothesis but not the other two (children and stranger typifications). Model fit was improved only slightly (adjusted R² increased from .159 to .162) with the addition of the societal image variables in Model 2. Several controls were significant. Those with greater educational attainment (p ≤ .10; β = −.089) endorsed the dispositional perception (i.e., sex offenders commit sex crime because of selfishness or immoral character) less so compared with individuals with lower levels of education. In comparison, conservatism (β = .200), religiosity (p ≤ .10; β = .089), the belief that sex crime has increased (β = .161), perceived community disorder (β = .105), and percent Republican (β = .102) were associated with a greater endorsement of the dispositional view of sex offending. Specifically, the three factors that exhibited the largest effects on the outcome variable were conservatism, the perception that sex offending has increased, and local disorder.

Table 4.

OLS Regression Analyses of the Effects of Societal Images on Assessment of the Causes of Sex Offending.

	Model 1			Model 2
Variables	b	RSE	St. Coef.	b	RSE	St. Coef.
Societal images
Perceived percent young	—	—	—	.002	(.002)	.040
Perceived percent female	—	—	—	.005^†	(.002)	.085
Perceived percent stranger	—	—	—	.000	(.002)	.007
Control variables
Female	−.068	(.081)	−.039	−.077	(.082)	−.044
Parent	−.013	(.084)	−.007	−.015	(.083)	−.008
Married	−.072	(.087)	−.040	−.068	(.089)	−.038
White	−.177^†	(.092)	−.087	−.145	(.092)	−.071
Age	.010	(.027)	.017	.008	(.027)	.013
Education	−.078*	(.039)	−.096	−.072^†	(.039)	−.089
Income	−.038	(.028)	−.069	−.039	(.028)	−.071
Conservatism	.184***	(.042)	.214	.172***	(.043)	.200
Religiosity	.041^†	(.023)	.087	.042^†	(.024)	.089
Sex crime increase	.161***	(.042)	.166	.156***	(.042)	.161
Perceived control	−.040	(.033)	−.048	−.051	(.033)	−.062
Sex crime victim	−.008	(.089)	−.004	−.003	(.088)	−.002
Community disorder	.118*	(.056)	.103	.122*	(.056)	.105
Percent younger than 18	−.000	(.012)	−.001	.001	(.012)	.003
Sqrt number of RSOs	−.010	(.013)	−.037	−.011	(.013)	−.040
Percent Republican	.006*	(.003)	.104	.006*	(.003)	.102
South	.030	(.076)	.016	.031	(.076)	.017
R ²		.188			.196
Adjusted R²		.159			.162

†

p < .10. *p < .05. **p < .01. ***p <.001 (two-tailed significance test).

Discussion

Overall, what can be gleaned from these findings? First, with the exception of the final set of models, the child victim typification strongly predicted both reformability and relative harm views. Female victims and stranger typifications were less predictive. Why? Females may be perceived as more culpable than children for their victimization. Indeed, “victim blaming” is primarily reserved for women rather than applied toward child victims of sex crime. To illustrate, Rogers and Davies (2007) explained, “children are generally considered to be sexually naïve,[hence] young children’s accounts of sexual events are deemed to be trustworthy, as they are seen as not likely to be capable of lying about sexual events” (p. 568). Accordingly, young children’s allegations about sexual experiences may be evaluated as more credible by the public than those reported by adult females. It follows that perhaps the origins of negative assessments of sex offenders (i.e., their reformability, damage caused by their offending) primarily flow from public concern about child (or “deserving”) victims rather than for adult females. Separately, recent research suggests that some Americans may not judge crime committed by strangers to present a unique danger to victims. To illustrate, Craun and Theriot (2009) reported that only 3 in 10 Americans assume a greater risk of sexual victimization by a stranger compared with an acquaintance, family member, or friend. It may also be that the public is equally repulsed by and unable to comprehend sexual offenses committed by acquaintances and strangers. Finally, it is not clear why the female victim typification—and even then, only weakly so—was linked with dispositional assessments of the causes of sex offending, but the other indicators were not. To be sure, very few studies have centered specifically on explaining public views about the factors that may drive sex offending and so little work exists to help place our findings into greater context. Having said that, it may be that perceptions related to the causes of sex crime represent distinct views from, say, judgments about offender reformability or relative harm to victims. As a result, a separate set of variables—beyond the ones examined here—may be more influential in shaping such perceptions.

Conclusion and Implications

Over six decades have passed since Edwin Sutherland penned his commentary criticizing “sexual psychopath” laws, or initial civil commitment statutes in the United States, and the faulty assumptions on which they rested. In contemporary society, the policies that apply to sex offenders go well beyond these measures. The current sex crime policy landscape emphasizes monitoring, control, and additional punishments for those convicted of sex offenses. Public opinion has been implicated as contributing to this “tough on sex crime” movement. Indeed, prior research has linked negative perceptions the public collectively holds regarding sex offenders to support for increasingly punitive laws (Pickett et al., 2013). Even so, left unaddressed by extant scholarship is an understanding of the origins of such perceptions. The current study that examined data from a large sample of Americans provides some answers. The most consistent finding surrounds child victim perceptions. In particular, believing that child victims are most at risk of sexual assault predicted negative judgments about sexual offenders and victimization—specifically concerning their amenability to be reformed, and the relative harm of sex crime. We theorize that the public is particularly likely to experience intuitive negative emotions in response to sex crimes committed against children and has difficulty comprehending such offenses. Thus, per the coherence hypothesis, the perception that sex offenders disproportionately target children cultivates a range of other negative and stigmatizing beliefs about sex offenders, as these views align with the public’s general conception of sex offenders as evil, monstrous “others.”

This brings us to the research implications of the study. Less consistent findings surrounded the origins of the public’s dispositional views of sex offending. The female typification was a marginally significant predictor of the judgment that immorality and selfishness cause sex offending. In this model, the child typification perception was not significant, nor was the stranger typification. It is not entirely clear what is driving this pattern. Virtually no studies have investigated predictors of public attributions of sex offending and so our results are difficult to place in context. One limitation of the current study is that the alpha value for the dispositional view index did not quite reach the conventional .70 threshold. Prior studies also generally report low alphas for causal attribution indices (see Pickett & Baker, 2014). Thus, future studies are needed that further assess the relationships observed herein using stronger, more reliable measures of assessments of the causes of sex offending.

In addition, prior public opinion studies have suggested that attribution styles regarding general and violent offending may be moderated by such factors as empathetic tendencies (Unnever & Cullen, 2009) and racial animus (Bridges & Steen, 1998). It is not clear if these predictors would also extend to views about the causes of sex offending. It is likely that a range of other factors, such as news consumption or fundamentalist religious beliefs (Grasmick & McGill, 1994), may predict attributions of sex offending (see generally, Unnever, Cochran, Cullen, & Applegate, 2010). Accordingly, additional studies that test specific theoretical propositions related to the causes of sex offending are warranted.

Separately, given the heterogeneity across sex offenders (e.g., Zilney & Zilney, 2009), there is a need for research that examines public perceptions of the various “types” of offenders. Although in step with recent scholarship (e.g., Button et al., 2013; Mancini, 2014), our study is limited in generalizing to specific sub-types of sex offenders as we measure attitudes concerning the generic “sex offender.” Another related shortcoming is that we include only one indicator of sex offender reformability. This is also consistent with previous work examining views about sex offender rehabilitation (Payne et al., 2010). Future work should develop alternative measures that tap attitudes—including views about reform—toward the different types of sex offenders (juvenile sex offenders, first-time sex offenders, non-contact sex offenders, repeat offenders with child victim preferences, etc.).

Ideally, future studies will work toward testing these theoretical propositions on a range of samples using a variety of research designs. In the current study, we relied on a Web-based sample. As discussed previously, despite the advantages of this approach over other methods, online surveys are not without their limitations. For instance, our sample is a probability sample of a nonprobability panel, and there were also some differences between the sample’s demographics and that of larger population. For these reasons, there is a possibility that our findings may not generalize beyond the opt-in panel. It follows that future studies should attempt to replicate our findings with nationally representative samples. It may also be pertinent to utilize qualitative methods to better understand nuances in public perceptions toward sex offenders (Fox, 2013), and also policymakers’ views, given their influence on the development of sex offender laws (see Meloy et al., 2013; Sample & Kadleck, 2008).

Because public opinion may have driven contemporary sex crime reforms (see generally, Burstein, 2009), determining the sources of public judgments of sex offending—particularly those that do not necessarily align with the reality of sex crime—is a critical first step toward understanding the tough on sex crime movement. Indeed, sex crime policymaking has become a “growth industry” in the United States (generally, Boyle, Ragusa-Salerno, Marcus, Passannante, & Furrer, 2014; Mancini, Barnes, & Mears, 2013). Nationally, the current sex offender population encompasses approximately 770,000 offenders (National Center for Missing and Exploited Children, 2013). Over a 5-year period, the number of offenders required to register in the United States increased by almost 20% (compare National Center for Missing and Exploited Children, 2008, 2013). Given the nearly universal adoption of “get tough” laws in recent years by all 50 states and the federal government (Mancini et al., 2013), the RSO population shows no signs of declining. To the extent that public opinion has initiated and shaped the development of laws and policies designed to manage this population, it is imperative that the origins of these views are identified and understood.

If subsequent research with representative samples finds evidence that public misperceptions of sex offenders exert an influence on policymaking, then efforts to dispel these misperceptions would be important. This could be accomplished through disseminating more accurate information about sexual victimization and risk prevention to the public. Under that backdrop, one approach advocated by Meloy and her colleagues (2013) is to follow the lead of medical researchers who have successfully marketed their scholarship as “newsworthy” (p. 449). Indeed, media and print outlets routinely have a “Health” section that provides brief summaries of recently published research in the medical fields. When applied to criminology, this approach—highlighting for instance, recent research examining crime and victimization, or crime prevention policies—may go far in providing a more accurate research base in which the public can draw on to form opinions about offending—including sex crime. To be sure, we are unlikely to witness dramatic gains in public knowledge overnight, but consistent dissemination over time may go far in gradually contributing to a more informed public debate about sex crime.

Footnotes

Declaration of Conflicting Interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) received no financial support for the research, authorship, and/or publication of this article.

Notes

Author Biographies

Christina Mancini, PhD, is an assistant professor at Virginia Commonwealth University’s L. Douglas Wilder School of Government and Public Affairs. She received her doctoral degree from Florida State University’s College of Criminology and Criminal Justice in 2009. She has published several articles in Criminology, Crime & Delinquency, and other crime and policy journals. She is currently involved in studies examining the emergence and efficacy of sex offender policy, public opinion, race and perceptions about offending, and violent victimization. Recently, she authored Sex Crime, Offenders, and Society: A Critical Look at Sexual Offending and Policy (Carolina Academic Press).

Justin T. Pickett, PhD, is an assistant professor in the School of Criminal Justice at the University at Albany, State University of New York. His research interests center broadly on public opinion about criminal justice. He is currently comparing different methods for measuring ambiguity in risk perceptions, and examining the role of intergroup contact as a determinant of Israeli Jews’ views about conciliatory solutions to the Israeli–Palestinian conflict.

References

Ackerman

A. R.

Sacks

Greenberg

D. F.

(2012). Legislation targeting sex offenders: Are recent policies effective in reducing rape? Justice Quarterly, 29, 858-887.

Anderson

A. L.

Sample

L. L.

Cain

C. M.

(2013). Residency restrictions for sex offenders: Public opinion on appropriate distances. Criminal Justice Policy Review. Doi:10.1177/0887403413513897.

Best

(1999). Random violence: How we talk about new crimes and new victims. Berkeley: University of California Press.

Bethlehem

Biffignandi

(2012). Handbook of Web surveys. Hoboken, NJ: Wiley.

Bhutta

C. B.

(2012). Not by the book: Facebook as a sampling frame. Sociological Methods & Research, 41, 57-88.

Blodorn

O’Brien

L. T.

(2013). Evaluations of White Americans versus Black Americans discrimination claimants’ political views and prejudicial attitudes. Journal of Experimental Social Psychology, 49, 211-216.

Borg

M. J.

(1997). The southern subculture of punitiveness? Regional variation in support for capital punishment. Journal of Research in Crime & Delinquency, 34, 25-45.

Boyle

D. J.

Ragusa-Salerno

L. M.

Marcus

A. F.

Passannante

M. R.

Furrer

(2014). Public knowledge and use of sexual offender Internet registries: Results from a random digit dialing telephone survey. Journal of Interpersonal Violence, 29, 1914-1932.

Bridges

G. S.

Steen

(1998). Racial disparities in official assessments of juvenile offenders: Attributional stereotypes as mediating mechanisms. American Sociological Review, 63, 554-570.

10.

Buhrmester

Kwang

Gosling

S.D.

(2011). Amazon’s Mechanical Turk: A new source of inexpensive, yet high-quality data? Perspectives on Psychological Science, 6, 3-5.

11.

Burstein

(2009). Public opinion, public policy, and democracy. In Leicht

Jenkins

J. C.

(Eds.), Handbook of politics: State and society in global perspective (pp. 63-79). New York, NY: Springer.

12.

Button

D. M.

Tewksbury

Mustaine

E. E.

Payne

B. K.

(2013). Factors contributing to perceptions about policies regarding the electronic monitoring of sex offenders: The role of demographic characteristics, victimization experiences, and social disorganization. International Journal of Offender Therapy and Comparative Criminology, 57, 25-54.

13.

Carmines

E. G.

Zeller

R. A.

(1979). Reliability and validity assessment. Thousand Oaks, CA: Sage.

14.

Chang

L. C.

Krosnick

J. A.

(2009). National surveys via RDD telephone versus the Internet: Comparing sample representativeness and response quality. Public Opinion Quarterly, 73, 641-678.

15.

Cortina

J. M.

(1993). What is coefficient alpha? An examination of theory and applications. Journal of Applied Psychology, 78, 98-104.

16.

Craun

S. W.

Theriot

M. T.

(2009). Misperceptions of sex offender perpetration considering the impact of sex offender registration. Journal of Interpersonal Violence, 24, 2057-2072.

17.

Finkelhor

Jones

L. M.

(2012). Have sexual abuse and physical abuse declined since the 1990s? Durham, NH: Crimes Against Children Research Center.

18.

Fox

K. J.

(2013). Incurable sex offenders, lousy judges, and the media: Moral panic sustenance in the age of new media. American Journal of Criminal Justice, 38, 160-181.

19.

Fuselier

D. A.

Durham

R. L.

Wurtele

S. K.

(2002). The child sexual abuser: Perceptions of college students and professionals. Sexual Abuse: A Journal of Research and Treatment, 14, 267-276.

20.

Garland

(2001). The culture of control: Crime and social order in contemporary society. Chicago, IL: University of Chicago Press.

21.

Garland

Sparks

(2000). Criminology, social theory, and the challenge of our times. British Journal of Criminology, 40, 189-204.

22.

Gosling

S. D.

Vazire

Srivastava

John

O. P.

(2004). Should we trust web-based studies? A comparative analysis of six preconceptions about Internet questionnaires. American Psychologist, 59, 93-104.

23.

Grasmick

H. G.

McGill

A. L.

(1994). Religion, attribution style, and punitiveness toward juvenile offenders. Criminology, 32, 23-46.

24.

Greer

(2007). News, media, victims and crime. In Davies

P. M.

Francis

Greer

(Eds.), Victims, crime, and society (pp. 20-49). London, England: Sage.

25.

Griswold

Wright

(2004). Cowbirds, locals, and the dynamic endurance of regionalism. American Journal of Sociology, 109, 1411-1451.

26.

Haidt

(2001). The emotional dog and its rational tail: A social intuitionist approach to moral judgment. Psychological Review, 108, 814-834.

27.

Haidt

Koller

S. H.

Dias

M. G.

(1993). Affect, culture, and morality, or is it wrong to eat your dog? Journal of Personality and Social Psychology, 65, 613-628.

28.

Harbeck

(2009). Shattered. Maitland, FL: Xulon Press.

29.

Jenkins

(1998). Moral panic: Changing concepts of the child molester in modern America. New Haven, CT: Yale University Press.

30.

Johnson

(2009). Anger about crime and support for punitive criminal justice policies. Punishment & Society, 11, 51-66.

31.

Kahneman

(2011). Thinking, fast and slow. New York, NY: Macmillan.

32.

Kano

Azuma

(2001). Use of SEM programs to precisely measure scale reliability. In Yanai

Okada

Shigemasu

Kano

Meulman

J. J.

(Eds.), New developments in psychometrics (pp. 141-148). Osaka, Japan: Springer.

33.

Kernsmith

P. D.

Craun

S. W.

Foster

(2009). Public attitudes toward sexual offenders and sex offender registration. Journal of Child Sexual Abuse, 18, 290-301.

34.

Kitzinger

(2004). Framing abuse: Media influence and public understanding of sexual violence against children. Ann Arbor, MI: Pluto Press.

35.

Levenson

J. S.

Brannon

Y. N.

Fortney

Baker

(2007). Public perceptions about sex offenders and community protection policies. Analyses of Social Issues and Public Policy, 7, 137-161.

36.

Lytle

(2013). Variation in criminal justice policy-making: An exploratory study using sex offender registration and community notification laws. Criminal Justice Policy Review. Doi: 10.1177/0887403413507274.

37.

Mancini

(2014). Examining factors that predict public concern about the collateral consequences of sex crime policy. Criminal Justice Policy Review, 25, 450-475.

38.

Mancini

Barnes

J. C.

Mears

D. P.

(2013). It varies from state to state: An examination of sex crime laws nationally. Criminal Justice Policy Review, 24, 166-198.

39.

Meloy

M. L.

Curtis

K. M.

Boatwright

(2013). The sponsors of sex offender bills speak up: Policymakers’ perceptions of sex offenders, sex crimes, and sex offender legislation. Criminal Justice and Behavior, 40, 438-452.

40.

National Center for Missing and Exploited Children. (2008). Map of registered sex offenders in the United States. Alexandria, VA: Author.

41.

National Center for Missing and Exploited Children. (2013). Map of registered sex offenders in the United States. Alexandria, VA: Author.

42.

Nhan

Polzer

Ferguson

(2012). “More dangerous than hitmen”: Judicial perceptions of sexual offenders. International Journal of Criminology and Sociological Theory, 5, 823-836.

43.

Nunnally

(1978). Psychometric theory. New York, NY: McGraw-Hill.

44.

Payne

B. K.

Tewksbury

Mustaine

E. E.

(2010). Attitudes about rehabilitating sex offenders: Demographic, victimization, and community-level influences. Journal of Criminal Justice, 38, 580-588.

45.

Pickett

J. T.

Baker

(2014). The pragmatic American: Empirical reality or methodological artifact? Criminology, 52, 195-222.

46.

Pickett

J. T.

Chiricos

(2012). Controlling other people’s children: Racialized views of delinquency and Whites’ punitive attitudes toward juvenile offenders. Criminology, 50, 673-710.

47.

Pickett

J. T.

Mancini

Mears

D. P.

(2013). Vulnerable victims, monstrous offenders, and unmanageable risk: Explaining public opinion on the social control of sex crime. Criminology, 51, 729-759.

48.

Planty

Langton

Krebs

Berzofsky

Smiley-McDonald

(2013). Female victims of sexual violence, 1994-2010. Washington, DC: Bureau of Justice Statistics.

49.

Rhodes

S. D.

Bowie

D. A.

Hergenrather

K. C.

(2003). Collecting behavioral data using the World Wide Web: Considerations for researchers. Journal of Epidemiology & Community Health, 57, 68-73.

50.

Rogers

Davies

(2007). Perceptions of victims and perpetrators in a depicted child sexual abuse case: Gender and age factors. Journal of Interpersonal Violence, 22, 566-584.

51.

Sample

L. L.

Kadleck

(2008). Sex offender laws: Legislators’ accounts of the need for policy. Criminal Justice Policy Review, 19, 40-62.

52.

Sampson

R. J.

Raudenbush

S. W.

(2004). Seeing disorder: Neighborhood stigma and the social construction of “broken windows.” Social Psychology Quarterly, 67, 319-342.

53.

Sanders

Clarke

H. D.

Stewart

M. C

Whiteley

(2007). Does mode matter for modeling political choice? Evidence from the 2005 British Election Study. Political Analysis, 15, 257-285.

54.

Schafran

L. H.

(1992). Maiming the soul: Judges, sentencing, and the myth of the nonviolent rapist. Fordham Urban Law Journal, 20, 439-453.

55.

Schmidt

W. C.

(1997). World Wide Web survey research: Benefits, potential problems, and solutions. Behavioral Research Methods, Instruments, & Computers, 29, 274-279.

56.

Sijtsma

(2009). On the use, the misuse, and the very limited usefulness of Cronbach’s alpha. Psychometrika, 74, 107-120.

57.

Skogan

W. G.

(1992). Disorder and decline: Crime and the spiral of decay in American neighborhoods. Berkley: University of California Press.

58.

Spencer

(2009). Sex offender as homo sacer. Punishment & Society, 11, 219-240.

59.

Sutherland

E. H.

(1950). The sexual psychopath laws. Journal of Criminal Law and Criminology, 40, 543-554.

60.

Tourangeau

Yan

(2007). Sensitive questions in surveys. Psychological Bulletin, 133, 859-883.

61.

Unnever

J. D.

Cochran

J. K.

Cullen

F. T.

Applegate

B. K.

(2010). The pragmatic American: Attributions of crime and the hydraulic relation hypothesis. Justice Quarterly, 27, 431-457.

62.

Unnever

J. D.

Cullen

F. T.

(2009). Empathetic identification and punitiveness: A middle-range theory of individual differences. Theoretical Criminology, 13, 283-312.

63.

Unnever

J. D.

Cullen

F. T.

Fisher

B. S.

(2007). “A liberal is someone who has not been mugged”: Criminal victimization and political beliefs. Justice Quarterly, 24, 309-334.

64.

U.S. Census Bureau. (2010). 2010 census national summary file of redistricting data. Washington, DC: Author.

65.

U.S. Census Bureau. (2012). State and county quick facts. Washington, DC: Author.

66.

Winship

Radbill

(1994). Sampling weights and regression analysis. Sociological Methods & Research, 23, 230-257.

67.

Zgoba

K. M.

(2004). Spin doctors and moral crusaders: The moral panic behind child safety legislation. Criminal Justice Studies, 17, 385-404.

68.

Zhang

(1999). Using the Internet for survey research: A case study. Journal of the American Society for Information Science, 51, 57-68.

69.

Zilney

L. J.

Zilney

L. A.

(2009). Perverts and predators: The making of sexual offending laws. Lanham, MD: Rowman & Littlefield.