Campus Criminal Victimization among Higher Education Students: A Diagnosis of Local Security in Porto

Abstract This study addressed and characterized direct and indirect criminal victimization among college students and examined the associations between victimization and other variables. The participants were 775 students of both genders with a mean age of 21.76 years. Data were collected through self-reports using the “Diagnosis of Local Security Questionnaire”. Overall, 8.6% of the students reported direct victimization, and 39.7% reported indirect victimization. The most reported crimes were robbery and theft, while the least prevalent were sexual offense, domestic violence, and fraud. Most incidents involved a stranger and occurred at night in the street. Direct victimization was associated with gender, age, marital status, student level, attendance status, and institution domain. There were significant associations between indirect victimization and nationality, student type, attendance status, and institution domain. The perception of (in)security was significantly associated with victimization. Therefore, physical and social measures are necessary to reduce crime and increase security.


Introduction
Since the 1970s, researchers have extensively studied victimization on higher education campuses, which is currently widely recognized not only as an educational concern but also as a social issue (Chekwa, Thomas, & Jones, 2013;Fox, Nobles, & Piquero, 2009;Hart & Colavito, 2011;Henson & Stone, 1999;Hibdon, Schafer, Lee, & Summers, 2016;Jacobsen, 2017;Jennings, Gover, & Pudrzynska, 2007;Merianos, King, & Vidourek, 2017;Tomsich, Gover, & Jennings, 2011;Volkwein, Szelest, & Lizotte, 1995). However, interest in crime and (in)security is not widespread but instead is mostly concentrated in Anglo-American countries, and it is risky to apply conclusions from studies in these countries to different cultures and contexts. This gap in the research on campus crime and (in)security was the starting point for the current research, which focused on an underresearched population, namely, Portuguese higher education students.
In the absence of previous studies on this topic, many doubts arise: Do Portuguese higher education students feel safe on campus? How many of these students are victims of campus crime? What characteristics and dynamics are involved in victimization incidents? Are sociodemographic variables related to victimization? Is there a relationship between victimization and the perception of (in)security? Overall, are Portuguese higher education students similar to or different from Anglo-American students? To address these research questions, we designed and implemented an empirical study that is presented here.
The paper begins with a literature review presenting the state of the research on campus victimization and (in)security. Next, the specific context of the study and its aims are described. The methodology section provides information on the data collection procedures, design, participants, and analysis plan. This is followed by the univariate and bivariate analyses, which are then discussed. Finally, a conclusion is offered, including implications, limitations and suggestions for future research.

Literature review
According to situational opportunity theories (recently reviewed by Wilcox & Cullen, 2018), namely, lifestyle-routine activities theory (Cohen, Kluegel, & Land, 1981) and a theory on situational crime prevention through environmental design (Newman, 1972), campuses can be hotspots for crime, and college students constitute an at-risk group for both direct and indirect victimization (Lee & Hilinski-Rosick, 2012). According to the former theory, the risk of victimization results from the interplay between time and space, involving the motivated offenders, target's and victim's proximity, target suitability, and lack of adequate guardianship; meanwhile, the latter theory stresses crime opportunities associated with site layout and design features (i.e., access control, target hardening, and surveillance potential). Campuses are free public spaces (including public transportation services) that are visited during the day and at night by a large transient population, especially young males, who are a high-offending group (e.g., Direc¸ão-Geral da Pol ıtica de Justiça, 2019; Hindelang, 1981). Additionally, college students, who are usually away from family and household members, own a large number of expensive items (e.g., mobile phones, laptops, and watches), tend to feel largely invulnerable to risk and thus engage in few protective measures and seem to be prone to alcohol and recreational drug consumption (Fisher & Wilkes, 2003;Lee & Hilinski-Rosick, 2012). Because campuses are potential hotspots for crime and because higher education students are potential victims, researchers have examined these topics intensively since the 1970s, with pioneering works from McPheters (1978) and Fox and Hellman (1985). To address the focus of our study, we reviewed mainly a set of empirical studies about campus criminal victimization (and its characteristics and dynamics) and perceptions of (in)security.
Although empirical data suggest that campuses are safer than the communities in which they are located (Baum & Klaus, 2005;Hart, 2003;Henson & Stone, 1999;Volkwein et al., 1995), taking a provocative approach, Jennings et al. (2007) asked, "Are institutions of higher learning safe?" Different parameters can be considered to address this question, but descriptive studies about victimization are usually a first step (Hart & Colavito, 2011), considering that a "strength of any victimization survey is not in its ability to examine rare crimes, but in its ability to obtain a better picture of very common minor crimes" (Henson & Stone, 1999, p.302). In the study by Jennings et al. (2007), 21.5% of students were direct victims, and 45.9% were indirect victims. In a study by Fox et al. (2009), direct victimization ranged from 14.19% for physical assault to 44.69% for property crime related to theft, while the rate of indirect victimization was 57.19%. Recently, Lee and Hilinski-Rosick (2012) conducted a study with 3472 college students and showed that 27% were direct victims and 38% were indirect victims of property crime. Moreover, regarding violent behaviors, 10% reported direct victimization, and 19% reported indirect victimization. This variation in the rates can be explained not only by actual differences but also by operational and methodological criteria, such as the assessed type of crime(s) or timeframe.
According to the aforementioned studies, campus victimization seems to be a frequent experience among higher education students. Therefore, in addition to providing answers to the question of "how many" students are being victimized, it has become essential to obtain deeper knowledge about victimization characteristics and dynamics (i.e., what, where, when, how, which, etc.), including the identities of the victims, incident variables, associated impacts and search for formal help. The available literature (Chekwa et al., 2013;Han, 2015;Hart & Colavito, 2011) suggests that burglary and theft/larceny are the most reported crimes, while rape and sexual assault are the least frequently reported. The role of the gender of the victim remains a controversial issue due to mixed results for both direct and indirect victimization. For instance, Baum and Klaus (2005) showed that direct victims were mainly males, while Han (2015) did not find gender differences. A similar pattern has been found for indirect victimization; while Tomsich et al. (2011) concluded that males reported higher rates of victimization than females, Fox et al. (2009) found no significant association between gender and indirect victimization. Data from the National Crime Victimization Survey (Baum & Klaus, 2005;Hart, 2003) showed that on-campus victimization occurred mainly during the day, while off-campus victimization was reported mainly at night. Moreover, 76% of the victims were not injured, and 25% reported physical damage. Only 34% of the victims reported the violent incidents to the police; when victims did report incidents, they were motivated mainly by preventive purposes (20%), such as an effort to stop the incident (19%) and to punish the offender (10%). When Henson and Stone (1999) asked participants about the main reasons for not reporting criminal victimization to the police, the most frequent answers were that the object was recovered or the offender was unsuccessful (27.2%), that the incident had been reported to another official (19.1%), and that the police would not want to be bothered (8.4%). Perceptions of (in)security have traditionally been analyzed together with criminal victimization experiences. For instance, using a five-point Likert scale, Merianos et al. (2017) asked students to rate how safe they felt on campus in general, and the authors found that the ratings were positive. Moreover, there were no differences when comparing students living on campus versus students living off campus, with ratings of 3.42 vs. 3.58, respectively. Maier and DePrince (2019) analyzed a sample of 133 students and concluded that 72% agreed or strongly agreed that they felt safe at the university, especially during the day. Overall, these findings suggest that students feel safe on campuses.
Although studies of campus victimization and (in)security are quite frequent in the USA, less is known about the European context, especially in southern Europe, including Portugal. This lack of scientific interest could be rooted, on the one hand, in traditional assumptions that college campuses are secure and safe places (Fisher et al., 1995;Fox et al., 2009;Jacobsen, 2017) and, on the other hand, in the absence of tragic events (e.g., shootings), which compromise current knowledge about the topic and make it difficult to develop proper social and institutional responses. Currently, Portuguese institutions do not have specific legislation or reporting practices, and there are no specific resources/programs or security policies to address campus victimization. Conversely, the media has recently paid increased attention to the topic, especially to violent episodes (e.g., Botelho, 2019;Pinto, 2014). To the best of our knowledge, only a single study addressed campus criminal victimization in Portugal; it focused on both students and staff from the University of Minho (Alves- Costa & Matos, 2014;Costa, 2011), and some of its conclusions were quite distinctive from those in the international literature. For instance, 10.3% of the participants were direct victims, while 14.8% reported indirect victimization. The most reported crimes were insults (5.2%), thievery (1.7%), and threats (1.6%), and 46.9% of the victims reported some kind of impact, mainly psychological (77.7%), followed by economic (33.3%) and physical (14.8%). Nonetheless, only 14.9% of the victims sought formal support.
This review of the research shows, on the one hand, the scientific developments achieved in the last decades and, on the other hand, the relevance of the topics. Nonetheless, it also indicates a geographic trend: most empirical evidence has come from Anglo-American countries, and there are differences among the findings from these countries. For instance, in a comparative study, Fisher and Wilkes (2003) concluded that English college students presented slightly higher rates of victimization than American students (37.5% vs. 36.1%, respectively). According to the Global Peace Index (2019), Portugal is the third most peaceful country in the world, while the United Kingdom (UK) and the United States of America occupy the 45 th and 128 th ranks, respectively. Nonetheless, based on the European Social Survey (European Comission, 2013), Portugal is in the top ten European countries with the most concerns about violent crime, while the UK is lower on the list. Consequently, simplistic generalizations to other populations should be avoided, and it is crucial to investigate other countries and cultures.

Current study
This study was based on an action research approach involving both the research team and the Metropolitan Police. Briefly, the study was conducted to collect evidence to be further applied in the development of a security plan. In the absence of previous studies, this research aimed to make a local security diagnosis (Direção-Geral de Administração Interna, 2009) by quantifying and describing criminal victimization reported by students on a Porto campus. The campus is open to the public and is located in the urban parish of Paranhos; it contains fourteen colleges/universities, both public and private, with different domains (e.g., social sciences, health, engineering, and economics) taught in undergraduate to doctoral programs. Therefore, the campus is visited by a large and heterogeneous population that is not restricted to students (who themselves represent a variety of demographic, geographic, and experiential backgrounds) or university staff but also includes inhabitants and passersby.
Porto is the second largest Portuguese city, and according to official data from the general community, it is also has second highest crime numbers. In 2018, 58765 incidents were reported to the police, while Lisbon (the capital) reported the highest numbers (n ¼ 87690). However, despite the national trend toward decreasing crime incidents (-2.6%), Porto showed an increase (0.7%). Differences among Lisbon and Porto are even more evident regarding violent crimes, which decreased by 14.8% in Lisbon but only by 2.7% in Porto. Overall, 15406 incidents were reported to the police in Porto in 2017, with 9290 being property crimes and 2896 being personal crimes (PORDATA, 2019). We cannot be sure if the campus is actually safer than the local community where it is located; thus, it is important to collect empirical evidence on the topic. To contribute to national and international knowledge about campus criminal victimization and to provide empirical foundations for social policy, this pioneer study addressed three aims: i) to quantify the frequency of direct and indirect victimization; ii) to describe and compare the characteristics and dynamics of direct and indirect victimization; and iii) to identify potential associations between sociodemographic variables, perception of (in)security and victimization.

Methodology
Next, we describe the methodological features of our study, which we structure under four topics. First, we present the data collection procedures, describing all the steps involved. Second, in the design subsection, we clarify the general characteristics and provide information about the assessed variables (and how they were measured and operationalized). We present a description of the study sample and selection criteria in the Participants subsection. Finally, in the fourth subsection -Analysis Planwe identify and justify the statistical methods.

Data collection procedures
After approval was obtained from the Internal Review Board, authorization for data collection was obtained for all fourteen previously identified universities/schools on the Porto campus. Next, students were invited, in groups or individually, to participate in a study about perceptions of (in)security and crime in the area where they studied. The study procedures (e.g., anonymity) and conditions of participation were presented, and written informed consent was obtained from all individuals who agreed to participate. Self-report surveys were gathered through a paper-and-pencil (administered in class) or online questionnaire (distributed via e-mail) according to the data collection strategy defined by each university/school. The measure took 20-30 minutes to complete. Participants did not receive any incentive to enroll in the study. Data were collected between 2015 and 2017.

Design
To address the study aims, a cross-sectional study was designed and performed, and all variables were quantitatively analyzed (although data were collected through qualitative/open-ended questions, as described later). Nonrandom sampling techniques were applied, and a convenience sample was formed.
Data were collected using the "Diagnosis of Local Security Questionnaire" (Sani & Nunes, 2013), a self-report measure that was specifically developed in collaboration with the Porto Metropolitan Police to evaluate the objective and subjective features of (in)security. The questionnaire has been validated in the Portuguese population and used intensively in different groups to perform local security audits (e.g., Sani & Nunes, 2012;Sani & Nunes, 2016, Sani & Nunes, 2017. The questionnaire has 67 closed-and open-ended questions, including skipped items, which allow for the collection of quantitative and qualitative data. The questions are grouped into five sections, including sociodemographic variables, perception of (in)security and fear of crime, direct and indirect victimization, social control and community participation. The current study focused on only three variables, namely, sociodemographic data, victimization, and perception of (in)security, which will be detailed below.
Concerning sociodemographic variables, participants were asked about their gender, age, nationality, marital status, education, university/college/school, and occupational status.
For direct victimization, often also labeled as personal victimization, one initial question asked, "In the last five years, have you been a victim of crime on campus?" ("yes" or "no" response). For those who answered affirmatively, additional information was collected, including the type of crime, damage (i.e., physical, psychological and/or property), time of day (i.e., day vs. night), space (i.e., house vs. street vs. other), whether they were alone vs. with someone else (an "I don't know" option was also available), and relationship with the offender (i.e., an acquaintance vs. a stranger vs. I don't know). Moreover, participants were asked whether they had contacted the police (and the reasons why or why not through an open-ended question) and formally registered the complaint. Measures adopted by the police as well as participants' levels of satisfaction with those measures were also assessed. For the indirect victimization variable, often also labeled as vicarious victimization, participants were asked, "In the last five years, has a friend, acquaintance or colleague been a victim of crime on campus?" ("yes" or "no" response). Similar to direct victimization, the same set of additional questions was asked to those who answered positively; however, a major distinction was made, namely, that instead of focusing on their own experiences, participants were asked to reference a criminal incident based on another person's experience (i.e., a friend, a colleague or an acquaintance). Therefore, for indirect victimization, participants were collateral informants, and an additional response option was added to all questions (i.e., "I don't know").
Perception of (in)security was assessed through a single question (i.e., "Do you feel you are studying on a safe campus?") with a dichotomous response scale (i.e., yes vs. no). Additionally, participants were further asked to explain their answers in an openended response. To minimize a potential question-order effect (Yang & Wyckoff, 2010), this item preceded the questions about victimization.

Participants
Individuals studying at the Porto campus were recruited as participants in this study. The Porto campus includes 14 schools/colleges/universities, 21.43% of which are private, and 78.57% of which are public. Only one inclusion criterion was used to select students, namely, that students must attend some higher education institution at the Asprela campus (geographical delimitation). Data were obtained from 778 students; three students were removed from the sample due to duplicate information or absence of answers.

Analysis plan
The quantitative data were analyzed using the IBM Statistical Package for Social Sciences software (IBM SPSS for Windows, version 25.0, IBM Corp, Armonk, NY, USA). According to our first and second aims and the descriptive, cross-sectional study design, univariate descriptive statistics were computed for all variables (i.e., sociodemographic data, (direct and indirect) victimization, and perception of (in)security).
Qualitative data from open-ended questions were initially coded through thematic analysis and then were further analyzed quantitatively. Moreover, to identify the potential associations between the sociodemographic variables, perception of (in)security and victimization (third aim), bivariate inferential statistical analyses, namely, chisquare and point-biserial correlation, were also performed.

Frequency of direct and indirect victimization
When asked whether they had been victims of any crime in the last five years, 8.6% (n ¼ 67) of the participants answered positively. Concerning indirect victimization, 39.7% (n ¼ 308) of the participants reported being aware that someone they knew had been a victim of crime on campus, namely, a friend, a colleague or an acquaintance (89.9%, n ¼ 275), a family member (3%, n ¼ 9), or an intimate partner (1.3%, n ¼ 4). Moreover, among those who reported victimization (n ¼ 334), 87.72% (n ¼ 293) had been victims of a single form of victimization, while 12.28% (n ¼ 41) reported being victims of both direct and indirect victimization. Table 1 presents the descriptive results of direct and indirect victimization; importantly, direct victims reported their own personal experiences, while indirect victims were collateral informants (i.e., they provided information about a criminal incident based on someone else's experience). Robbery or attempted robbery and theft were the most reported types of crimes. Fraud, sexual offense, and intimidation were less reported crimes by direct victims, while road traffic crime, discrimination, and domestic violence were less reported by indirect victims. For both types of victimization, individuals stated that the crimes mainly involved a stranger and occurred at night in the street. Of the other places where crimes were reported to have occurred, higher education institutions and public transportation were the main areas reported. A lower percentage of direct victims than indirect victims reported themselves/the victim being alone when the crime occurred. A high percentage of the individuals who reported indirect victimization indicated that they were not aware whether other people were present when the crime occurred.

Characteristics and dynamics of direct and indirect victimization
Almost all direct victims reported some kind of damage associated with the crime; however, the damage values were lower for indirect victimization than for direct victimization. In both cases, property damage and psychological damage had high frequencies.
A higher number of direct victims than indirect victims reported submitting a formal complaint (29.2% n ¼ 90 vs. 49.3% n ¼ 33, respectively). The main motivations for submitting formal complaints included to report property recovery/damage (direct victimization ¼ 21.9%, n ¼ 7 vs. indirect victimization ¼ 7.9%, n ¼ 7), to inform the police/seek justice (direct victimization ¼ 12.6%, n ¼ 4 vs. indirect victimization ¼ 6.8%, n ¼ 6), and to fulfill one's civil duty/provide statistical information (direct victimization ¼ 12.5%, n ¼ 4 vs. indirect victimization ¼ 6.8%, n ¼ 6). Conversely, worthlessness of submitting a complaint/untrustworthiness of authorities (direct victimization ¼ 58%, n ¼ 18 vs. indirect victimization ¼ 7.3%, n ¼ 3), absence of proof/ inability to identify the offender (direct victimization ¼ 9.7%, n ¼ 3 vs. indirect victimization ¼ 2.4%, n ¼ 1), lack of necessity (direct victimization ¼ 9.6%, n ¼ 3 vs. indirect victimization ¼ 9.7%, n ¼ 4) due to the recovery of belongings or offender arrest, and other reasons (e.g., civil duty, high frequency of the crime, and insurance issues) were the main reasons presented by those who did not complain. The majority of the participants did not know or did not specify the measures adopted by the police. Among those who were aware of those measures, they most frequently indicated that the police had filed a formal complaint (direct victimization ¼ 16.1%, n ¼ 5 vs. indirect victimization ¼ 9.2%, n ¼ 7), contacted them for statement (for direct victimization) (6.5%, n ¼ 2) and attempted to arrest the offender (for indirect victimization) (7.9%, n ¼ 6). More than three-quarters of the direct victims (75.9%, n ¼ 22) mentioned that they were not satisfied with police intervention, while for those participants that reported indirect victimization, most were not aware of whether the victims were satisfied (or not) with police intervention (57.3%, n ¼ 59) or were not satisfied (28.2%, n ¼ 29). Ineffectiveness (direct victimization ¼ 50.0%, n ¼ 11 vs. indirect victimization ¼ 69.2%, n ¼ 18) was the main reason mentioned by those who were not satisfied with police intervention, and effectiveness (direct victimization ¼ 83.4%, n ¼ 5 vs. indirect victimization ¼ 26.7%, n ¼ 4) was the reason presented by those who were satisfied.
When asked whether they felt safe on campus, the majority of participants (73.4%, n ¼ 569) answered positively, explaining that their choice was mainly based on prior experience/observation (47.6%, n ¼ 369). In contrast, more than a quarter of the participants (26.6%, n ¼ 206) reported feelings of insecurity, with prior experience/observation being the reason presented by 3.4% (n ¼ 26) of individuals. The chi-square test revealed a significant association between perception of (in)security and victimization (v 2 (1) ¼ 37.59, p < .001 for direct victimization and v 2 (1) ¼ 90.12, p < .001 for indirect victimization). More specifically, victims, especially direct victims, reported more feelings of insecurity.

Discussion
Despite cultural differences, there are several similarities between our findings and findings from international studies, which will be presented and discussed next.
Although the majority of students evaluated the campus as a safe place, similar to findings from a study by Maier and DePrince (2019), one-quarter of students reported feeling insecure, mainly due to past experience. Indeed, nearly 9% of students selfidentified as direct victims, and 40% reported experiencing indirect victimization. The percentage of individuals who experienced direct victimization is in line with the percentages reported in Alves- Costa and Matos (2014), Cook andFox (2011) andTomsich et al. (2011). In contrast, the percentage of those who experienced indirect victimization is slightly higher than the percentages reported in other studies, and this difference can be explained by the comprehensiveness of the potential targets included in the current study (e.g., friend, colleague, acquaintance). According to our findings, direct victims tended to be male, older, and full-time students who were pursuing a bachelor's degree, which is consistent with previous studies. For instance, Hart and Miethe (2011) found that in the campus context, males represent the majority of victims and offenders. Additionally, in a study about student characteristics, Han (2015) concluded that attendance status, age, program completion, residence status, and grant recipient predicted campus victimization. Surprisingly, only one variablethe institutional domainwas associated with both direct and indirect victimization; more specifically, victims tended to attend engineering universities/schools. This finding can be understood not only by the proportion of those students in our sample but also by contextual variables (e.g., the predominance of male students). Based on our findings on direct and indirect victimization, most crimes involved a stranger as the offender, occurred at night, occurred in the street or at the university, and occurred when the victim was alone. These conclusions are consistent with previous results (e.g., Costa, 2011;Hart & Miethe, 2011) and can be seen as evidence supporting situational opportunity theories (Wilcox & Cullen, 2018). Moreover, the Porto campus layout and design may also influence the incidence of crime. For instance, based on situational crime prevention through environmental design principles (Newman, 1972), we hypothesized that the extension of the campus and the permeability of the campus border may compromise access control and target hardening, while the absence of cameras and poor lighting zones may jeopardize surveillance potential.
Our results also showed that robbery and theft were the most reported types of crime, corroborating not only international data (Baum & Klaus, 2005;Jennings et al., 2007) but also national community-level trends (PORDATA., 2019). Nonetheless, our results contradict those from Alves- Costa and Matos (2014), who found that personal crimes were most prevalent. Since both studies focused on higher education campuses, this discrepancy is quite interesting and is probably explained by methodological and cultural reasons. Indeed, while we assessed victimization through a general question based on self-definition, Alves- Costa and Matos (2014) asked about personal experiences across several types of crime. The authors observed high values for behaviors such as insults, threats, and coercion, which are quite common in initiation rituals (hazing) of new students in Portugal and are not generally perceived as crimes. None of our participants reported being victims of direct sexual offenses, and 2.9% of the individuals reported indirect victimization by sexual offenses, suggesting sexual offenses are rare. Although a similar conclusion was reached by other authors (e.g., Chekwa et al., 2013;Jennings et al., 2007), based on the national data from the Associac¸ão Portuguesa de Apoio a V ıtima (APAV) (2018), it is unclear if those findings are truly representative or if sexual offenses were underreported, for instance, due to secrecy issues (Fisher et al., 1995). Notwithstanding, Bosick, Rennison, Gover, and Dodge (2012) studied police reporting through the life course and concluded that reports of sexual abuse decrease significantly between the ages of 16 and 24 years old. The frequency of domestic violence was also very low in our study, in contrast to other studies (Fox et al., 2009;Machado, Caridade, & Martins, 2010;Mengo & Black, 2016). This discordance can most likely be explained by the fact that we focused only on campus victimization, while domestic violence tends to occur in other more private contexts (e.g., the home).
Last, similar to the studies by Baum and Klaus (2005) and Hart and Colavito (2011), our study showed that the majority of the direct victims did not seek police support, and only thirty students presented a formal complaint. Moreover, 75.6% of the direct victims were unsatisfied with police intervention. Based on the results from Bosick et al. (2012), the percentage of violence reported to the police differs by victim age; more specifically, the percentage tends to be higher in older participants. Regarding the reasons for not reporting criminal victimization to the police, we determined that 75% of participants thought that doing so would be worthless or that the police were untrustworthy, whereas Henson and Stone (1999) found that only 1.3% of the participants considered the police to be inefficient, ineffective or biased. Based on the data from the European Social Survey (European Comission, 2013), this lack of trust in police action may be rooted in cultural idiosyncrasies; for instance, the relation between fear of crime and quality of life seems to be particularly strong in countries of southern Europe, and "those people reporting that fear of crime reduces their quality of life have less trust in the police and are less likely to think that the police are doing a good job" (p.19).

Conclusion
Research about campus criminal victimization and the perception of safety has remained limited to Anglo-American countries and has ignored some specific features (e.g., campus characteristics, socioeconomic variables, and crime rates); therefore, the current study aimed to address this gap in the literature by providing data about a southern European country. This study also contributes to the available literature since Portugal has been negligent regarding campus criminal victimization: there are no official data, no reporting practices, and no legal background or victimization programs. Moreover, there is no specific police division dedicated to campus safety, and overall policing practices remain traditional. Despite this status quo, we concluded that campus victimization and the perception of (in)security are also relevant issues for students attending Portuguese higher education institutions. Indeed, the campus environment is where students attend classes, study, live, and socialize. Based on our results, it is also a place where they become victims of crime.
Notably, being a crime victim seems to affect quality of life and well-being (Hanson, Sawyer, Genelle, Begle, & Hubel, 2010;Tan & Haining, 2016), as well as grade point average and the school dropout rate (Mengo & Black, 2016). Moreover, 70% of participants argued that security was a key factor for choosing a particular college or university (Chekwa et al., 2013).
In the absence of a panacea to eliminate campus victimization (Su et al., 2013), attention and (re)action are urgently required. Indeed, this study offers several implications for promoting security practices and policies that should be considered by the different stakeholders involved in campus security (Fox et al., 2009;Hart & Colavito, 2011;Volkwein et al., 1995). First, as recommended by Jennings et al. (2007) and Merianos et al. (2017), both physical and social changes at the individual and community levels should be designed and applied to create safer learning environments. To the best of our knowledge, no measures have been assessed or applied in Portugal; therefore, to avoid the risks of a one-size-fits-all solution, measures should be carefully analyzed, discussed, prioritized, and implemented according to this particular community (Fisher & Wilkes, 2003;Flannery & Quinn-Leering, 2000). As this study was a descriptive study, it seems reasonable to establish a work group or task force including administrators, police, professors, faculty, students, parents, politicians, and researchers to further assess and prevent campus criminal victimization. Additionally, our findings may also be informative for the development of awareness, educational and prevention campaigns or programs, for instance, those covering specific topics (e.g., domestic violence, victim rights) and targets (e.g., male victims). Notably, psychological damage was reported by more than 30% of the victims, suggesting that special services could be developed to provide support. Procedures and policies applied to decrease the incidence of crime and increase perceptions of security should be carefully followed to collect empirical evidence about their efficacy.
This study has several limitations. First, it relied on retrospective information collected through a cross-sectional design. Consequently, no causal relationships can be established, and memory bias may have influenced our findings. Moreover, the data were collected through self-reports, which can be biased by social desirability. In the current study, victims were identified through self-definition, which could be an additional limitation. Indeed, the use of a self-definition strategy requires individuals to not only be aware of which behaviors constitute crime but also recognize themselves as victims. Additionally, selection bias may also have interfered with our results. This study focused solely on criminal victimization within the context of the Porto campus and a convenience sample of students; as a result, caution should be taken when generalizing our results. Although the students represent an important group of users at higher education campuses, their findings and patterns should not be applied to other groups (e.g., faculty, professors, and, in this particular case, passersby).
Further work needs to be carried out to investigate student variables (e.g., lifestyle), to assess physical spaces and to characterize environments. Second, to monitor crime and security perceptions, it is important to gather data systematically from not only students but also faculty and staff. This type of research could provide knowledge about the rates and trends of the occurrence of crime. Third, we focused on "victims" of a diversified set of crimes, which allowed a broad overview but disregarded specificities. Therefore, future studies should include comparative analyses (e.g., personal crimes vs. property crimes). In our sample, 5.3% of the participants reported both direct and indirect victimization; it is reasonable to suspect that these victims may represent a special group deserving further analysis.

Disclosure statement
The authors declare no conflict of interest.

Funding
This work was financed by National Funds through FCT (Foundation for Science and Technology) under the project "LookCrim j Looking at crime: Communities and physical spaces" (PTDC/DIR-DCP/28120/2017).