Fear of Crime in South Korea

This study provides analyses of data on crime‐associated trepidation obtained from surveys administered to college students in South Korea. The survey contained questions about, and the analyses distinguished between, offense‐specific fears (fear of burglary and fear of home invasion), perceived risk of victimization (day and night), and crime avoidance behaviors (avoidance of nocturnal activity and avoidance of particular areas). Regression analyses of the data show that victimization was not consistently associated with crime‐associated trepidation, while gender significantly impacted all measures of concern about crime. Women were more likely than men to report being fearful, perceiving risk, and crime avoidance behaviors. Building upon prior scholarship (for example, Madriz 1997; Stanko 1989) and considering the social context in which the data were gathered, it is herein suggested that the gendered variation in crime‐associated anxiety may reflect patriarchal power relations. The methodological and policy implications of the study are also discussed.


Introduction
Although the damage which results from fear of crime is not as easily observable as the damage which results from corporeal crimes such as assault and burglary, trepidation of crime has palpable consequences. Research suggests people may avoid specific neighborhoods owing to fear of crime and, as a consequence, the quality of life in the area may deteriorate. Business owners may lose clients, lose quality employees, and find it difficult to attract new clients and hire new employees in an area perceived as dangerous. Consequently, such businesses may fail or relocate to other areas, which means fewer opportunities and services for residents of neighborhoods perceived to be dangerous. Research also indicates that fear of crime may degrade the quality of communal life by discouraging social interaction, with residents of neighborhoods perceived as dangerous being hesitant to spend time outside or interact with neighbors (Fisher 1991;Foster and Giles-Corti 2008;Lane 2002;Lane and Meeker 2004;Liska, Sanchirico and Reed 1988;Liska and Warner 1991).
Fear of crime may have an especially detrimental impact on the quality of life among women. Research has indicated women may refrain from consuming alcohol at social gatherings, flirting with men, socially interacting with men, and even going out at night owing to concerns about personal safety (Hickman and Muehlenhard 1997;McKibbin et al. 2011;May, Rader and Goodrum 2010;Ngo and Paternoster 2013;Warr 1985;Young 1992). While most studies which suggested fear constricts women's autonomy were conducted in the western hemisphere, Sur's (2014: 224) qualitative research conducted in Kolkata, India also indicated that fear of crime had an astringent impact 'on women's everyday lives as women engage in self-regulation and self-policing while negotiating risks of harm in public'.
In brief, fear of crime may have multiple effects such as limiting autonomy, constraining social interaction, and hindering business opportunities. There is even evidence that 'fear of crime is significantly associated with poor health' (Chandola 2001: 113). Although there is a sizeable body of literature on fear of crime, scant research has been conducted in Asia. This study contributes to the limited literature on fear of crime in Asia.

Literature review
There is substantial literature on concerns about crime. Researchers have examined the impact of numerous variables such as age, gender, race/ethnicity, perceived disorder, and victimization on crime-associated anxieties (Eitle and Taylor 2008;Hale 1996;Tseloni and Zarafonitou 2008). However, owing to the similarities in age and race/ethnicity among the respondents in this study and the limitations of the data, the research on the impact of several of the aforementioned variables on fear of crime is of no immediate relevance. This literature review focuses on gender and victimization as they are the only variables pertinent to this work.
The most consistent finding in the literature is that concerns about crime tend to be greater among women than among men. The tendency for crime-associated trepidation among women to be greater than that among men has been documented in numerous nations, inclusive of: Australia (Australian Bureau of Statistics 2010), Botswana (Johnson 2006), Canada (Alvi et al. 2001;Weinrath and Gartrell 1996), China Zhang et al. 2009), England (Gray, Jackson andFarrall 2011), Ghana (Adu-Mireku 2002), Greece (Tseloni and Zarafonitou 2008), Iceland (Kristjansson 2007), India (Chockalingam and Srinivasan 2009), Japan (Chockalingam and Srinivasan 2009), Mexico (Vilalta 2011(Vilalta , 2012, Scotland (Kristjansson 2007), Sweden (Özacilar 2013;Smith and Torstensson 1997), Turkey (Karakus, McGarrell and Basibuyuk 2010), and the United States (Clemente and Kleiman 1977;Eitle and Taylor 2008;Haynie 1998;Lane, Gover and Dahod 2009). One prior study conducted in South Korea (Lee 1997) also suggested fear of crime was problematic among women, but the study focused exclusively on women and did not contain an analysis of gendered variation in fear. IJCJ&SD 118 Online version via www.crimejusticejournal.com © 2016 5(4) As is inevitable, there are a few discrepancies. Nalla, Johnson, and Hayes-Smith's (2011) analyses of data from India showed that women in urban environments were more fearful of crime than were men but, in rural areas, gender was not significantly associated with fear of crime. A couple of studies conducted in the United States indicated that women were significantly more likely than men to be fearful of violent victimization, but gender did not significantly impact concerns about non-violent property offenses (Rountree and Land 1996;Schafer, Huebner and Bynum 2006). An examination of British Crime Survey data indicated that women were more fearful than men of violent crime, while men were more fearful than women of property crime (Moore and Shepherd 2007). A study of fear of gang crimes among Hispanic high school students in Texas (USA) showed the opposite: Namely, that female students were more fearful than were male students of gang-related theft but gender did not significantly impact fear of gang-related assault (Brown and Benedict 2009).
As to the effects of victimization on fear of crime, there is no consensus in the literature. Whereas analyses of data from China ), England (Gray, Jackson and Farrall 2011), Greece (Tseloni and Zarafonitou 2008) and Scotland (Kristjansson 2007) indicated victimization experiences were significantly associated with heightened concerns about crime, analyses of data from Ghana (Adu-Mireku 2002), Iceland (Kristjansson 2007), Mexico (Vilalta 2012) and the United States (Cook and Fox 2011) indicated victimization experiences did not significantly affect such concerns. Contrary to the commonsensical notion that violent victimization should have a more profound impact than non-violent victimization on concerns about crime, several studies suggested the opposite. An assessment of national survey data from the United States (Toseland 1982) and an examination of survey data from Tianjin, China ) showed that property victimization (for example, burglary) impacted fear of crime, but violent victimization did not. Smith and Hill's (1991: 232) analyses of data from North Carolina (USA) indicated that property crime victimization or a combination of property crime and personal crime victimization was associated with heightened fear of crime, while 'personal victimization alone turns out to have no measurable effect on levels of expressed fear'. Johnson's (2006) analyses of data from Botswana showed that burglary victimization, fraud victimization, and sex crime victimization had a significant impact on fear of crime; robbery victimization did not significantly affect fear of crime; and vehicular theft victimization significantly reduced fear of crime. Nalla and colleagues' (2011) analyses of data from India showed that violent victimization impacted fear of crime among residents of urban areas but not residents of rural areas. Chockalingam and Srinivasan's (2009) research suggested victimization had no impact on fear of crime among college students in India and was inversely associated with fear among college students in Japan. Winkel's (1998) analyses of longitudinal data from the Netherlands showed that crime victims expressed enhanced concern about crime immediately after the victimization incident but, as they recovered from the incident, they gained confidence in their abilities to cope with crime and became less concerned about victimization than persons who had not been victimized. Weinrath and Gartrell's (1996) analysis of data from Edmonton, Alberta (Canada) indicated that assault victimization was associated with heightened fear of crime among young adults, but reduced fear of crime among the elderly.
In brief, most research has indicated that fear of crime tends to be greater among women than among men. In contrast, there is no consensus on the impact of victimization or the manner in which the effects of victimization on fear may be mediated by other variables such as age and gender.

Research setting
South Korea is a republic with democratically elected leaders and a population of roughly 50 million people. South Korea is a member of the G-20 and the Organisation for Economic Concisely summarized, South Korea is a rapidly developing and reasonably safe nation with a highly urban populace. Home to almost half of the South Korean citizenry, the Seoul-Gyeonggi Province metropolitan area is a densely populated and dynamic global city. The present data were obtained via surveys administered to college students in the Seoul-Gyeonggi Province metropolitan area.

Sample
The college where the data were gathered was selected as a matter of convenience, not randomly chosen from the multitude of institutions of higher education in the Seoul-Gyeonggi Province metropolitan area. The college was non-residential, attended by students from throughout the Seoul-Gyeonggi Province area, with the majority of students (approximately 85 per cent) being women. The gender composition of the student body reflected the facts that the college was founded as a nursing school, that nursing was the largest program at the college, and that several other fields of study at the college (for example, office administration, early childhood education) are in occupational fields with predominantly women practitioners. Because of the student demographics, a purposive cluster sampling technique was employed.
Initially, first year nursing students were surveyed. As nursing is the largest program at the college, targeting nursing majors afforded the best opportunity to achieve a sizeable sample. The focus on first year nursing students was based on the accessibility of the classes. As the nursing students advance, they spend increasing time in hospitals engaged in practical exercises. Thus, first year students were most accessible. Next, majors popular among men (for example, Internet information) were purposively targeted in an effort to obtain a reasonable level of male involvement. In order to avoid any artificially generated age discrepancies between the genders, freshman-level classes in subjects popular among men were targeted.
IJCJ&SD 120 Online version via www.crimejusticejournal.com © 2016 5(4) The aforementioned strategies yielded a sample of 303 respondents (see Table 1). Despite efforts to obtain a sizeable number of men, almost two thirds of the respondents were women (65 per cent). The students surveyed were, on average, around 21 years of age (Mean = 20.78). For readers unfamiliar with Korean culture, it must be noted that South Koreans have a distinctive method of calculating age, with the period of gestation (treated as roughly a year) being included when determining age. The average reported age of the respondents is therefore likely a year greater than would be the case if South Koreans utilized the more common method of calculating age using date of birth as the starting point. While the use of a non-probability sample is not ideal, the challenges associated with the procurement of a random sample from one of the largest urban centers on the plant -a megalopolis composed of numerous municipalities sprawled across a couple of provinces -are considerable and well beyond the resources which were available for the project from which the present data were drawn. Moreover, the use of a non-probability sample in studies of this nature is not unusual. Several studies of fear of crime conducted in Asia and elsewhere in the world have utilized non-random samples, inclusive of non-probability samples of secondary and tertiary students (for example, Brown and Benedict 2009;Chockalingam and Srinivasan 2009;Lane, Gover and Dahod 2009;Pryor and Hughes 2013;Zhong 2010).

Survey design
The survey project from which the present data were culled was of a general nature and included queries pertinent to numerous topics (for example, fear of crime, confidence in the police) but, because of concerns about issues such as the time required to complete the surveys, the number of questions per topic was kept to a minimum. As for victimization, respondents were asked whether they had ever been the victim of a burglary, a non-violent theft (had their purse or wallet stolen), or a physical assault. The survey also contained questions pertinent to trepidation of crime which were modeled upon extant conceptual frameworks and methodological strategies.
Although a sizeable body of literature on trepidation of crime exists, there is no consensus as to the best means of conceptualizing and gauging such concerns (Farrall et al. 1997;Hale 1996;Tseloni and Zarafonitou 2008). While no single measure or set of measures of concerns about crime has been universally accepted, numerous researchers distinguished between 'fear' of crime and 'perceived risk' of victimization. Fear has been conceptualized as an emotional concern about victimization while perceived risk is conceptualized as a mental assessment of the likelihood of victimization (Cook and Fox 2011;May, Rader and Goodrum 2010;Warr 1987). Inquiries about specific forms of crime such as asking whether a person is concerned about being the victim of a burglary have been treated as measures of 'fear' of victimization while general queries about safety such as asking whether a person is afraid to walk around the neighborhood after dark have been treated as measures of 'perceived risk' of victimization (Ferraro and LaGrange 1987;May 2001;Rountree and Land 1996;Warr and Stafford 1983).
Most studies which documented a distinction between fear and perceived risk of victimization were conducted in the western hemisphere. The data analyzed herein permit assessment of whether Asians' concerns about crime are similarly bifurcated. Consistent with Rountree and Land's (1996: 1355-1356) treatment of 'Offense-Specific Fear of Crime,' respondents were asked whether they were fearful someone would break into their home when they were away (fear of burglary) and whether they were fearful someone would break into their home when they were there (fear of home invasion). As for perceived risk of victimization, respondents were asked if they were concerned about their safety when away from home during the day and if they were concerned about their safety when away from home at night. Such measures of anxiety about crime have been previously used in studies conducted on several different continents (Kanan and Pruitt 2002;Karakus, McGarrell and Basibuyuk 2010;Kristjansson 2007), including Asia Zhong 2010).
Based on research which suggested anxieties about crime can affect behavior (Hickman and Muehlenhard 1997;Lane, Gover and Dahod 2009;Liska, Sanchirico and Reed 1988), the survey also included questions about crime avoidance behavior. Questions about crime avoidance have not been as frequently used as questions about fear and perceived risk (May, Rader and Goodrum 2010) but the inclusion of such questions was considered important as the measures go beyond abstract concerns about crime and serve as indicators of the extent to which crimeassociated trepidation affects quality of life. In other words, it is one thing to say 'I am concerned about crime when I go out at night' and it is something entirely different to say 'I am so concerned about crime that I do not go out at night'.
Dichotomous response options (Yes / No) were used for the questions about fear of crime, perceived risk of victimization, crime avoidance behaviors, and victimization (see Table 1). The use of specific questions with limited response options is common when conducting research with small samples in developing nations. For instance, Adu-Mireku (2002) assessed fear of crime in Ghana utilizing a single question about fear with dichotomous response categories. In a study of fear of crime among women in Korea, Lee (1997: 321) relied upon two questions pertinent to fear and utilized binary measures which distinguished solely between 'whether the respondents felt safe or unsafe'. Some researchers who tried assessing fear of crime in Asia with multinomial measures encountered difficulties with the data (Chockalingam and Srinivasan 2009). To provide a couple of specifics, Zhang and colleagues (2009) utilized multinomial response options in their study of fear of crime in China, as did Nalla and colleagues (2011) in their study of fear of crime in India but, in both studies, the data obtained were so skewed that the multinomial measures had to be collapsed into dichotomous measures for statistical analysis.

Data analysis and findings
Descriptive analyses of the data (see Table 2) indicate that, except for fear of burglary, women were more than twice as likely as men to report fear, perceived risk, and avoidance behavior. Almost half of the women reported avoiding nocturnal ventures whereas less than a tenth of the men reported such behaviors. The analyses also indicate the impact of victimization on concerns about crime may vary depending on the type of victimization and the measure of concern about crime. Respondents who reported having been a burglary victim were more likely than those who reported no such victimization to have reported fear of home invasion, fear of burglary, concern about their safety during the day, and concern about their safety at night, but the relationship between burglary victimization and crime avoidance behaviors was IJCJ&SD 122 Online version via www.crimejusticejournal.com © 2016 5(4) inverse. The relationship between theft victimization and fear, perceived risk, and avoidance behaviors was similarly inconsistent. Physical assault victimization had little impact on concerns about crime, but the effects were directionally consistent. Assault victims reported less fear, less perceived risk, and less avoidance behavior than respondents who had not experienced an assault. Logistic regression analyses were conducted to determine whether gender or victimization significantly influence trepidation of crime. As shown in Table 3, gender significantly impacted all measures of concerns about crime. Burglary victimization significantly impacted fear of home invasion, fear of burglary, and perceived risk of victimization during the day but, in every other case, the relationship between victimization and concerns about crime was insignificant. 3 As indicated by the confidence levels, the impact of gender on the various measures of concern about crime was greater than the impact of burglary victimization. While logistic regression is appropriate for analyses of dichotomous variables (DeMaris 1995;Norusis 2005), the pseudo R 2 values in logistic regression models do not indicate the level of explained variance as do the R 2 values in linear regression models (Hu, Shao and Palta 2006;Long 1997). It is thus unclear how much of the variance can be attributed to gender and victimization.

Discussion
The present findings are consistent with the extant literature in a few ways. First, the finding that burglary was the only form of victimization significantly associated with concerns about victimization -and that burglary victimization was correlated with fear of burglary, fear of IJCJ&SD 123 Online version via www.crimejusticejournal.com © 2016 5(4) home invasion, and diurnal perceived risk of victimization, but not nocturnal perceived risk of victimization or either type of crime avoidance behavior -are concordant with the literature which suggests victimization does not consistently impact fear of crime (Adu-Mireku 2002; Gray, Jackson and Farrall 2011;Nalla, Johnson and Hayes-Smith 2011). Second, consistent with research conducted in the western hemisphere which documented a bifurcation between fear of crime and perceived risk of victimization, and differences in the variables which affect these forms of crime-associated trepidation (Kanan and Pruitt 2002;Lane and Meeker 2010;Rountree and Land 1996), the findings that the respondents reported different levels of fear and perceived risk, and that gender and victimization had varying degrees of impact on the different measures, suggests that crime-associated trepidation in East Asia may function similarly to such trepidation elsewhere in the world. Most prior research on fear of crime in Asia utilized data obtained from only one or two questions pertinent to fear (Lee 1997;Liu et al. 2009;Zhong 2010). By means of analyzing data obtained from multiple measures of fear, perceived risk, and crime avoidance, this study expands the literature on crime-associated concern in Asia and indicates that future researchers who study fear of crime in Asia may be well served by employing a multidimensional conceptualization of crime-associated anxieties.
Finally, gender was significantly associated with all measures of crime-associated trepidation. This finding adds credence to the body of international literature which suggests the tendency for women to be more fearful of crime than are men is a global norm (for example, Gray, Jackson and Farrall 2011;Johnson 2006;Kristjansson 2007). Lee's (1997: 321-322) study of fear of crime in South Korea showed that the majority of women surveyed (75.5%) reported being fearful of crime while walking alone at night, but Lee only surveyed women and was unable to examine the impact of gender on fear of crime among South Koreans. The present analyses provide evidence of the extent of gendered variance in fear, perceived risk, and crime avoidance behaviors in South Korea. Of special note was the impact of gender on avoidance behaviors, with roughly half of the women having reported avoidance of specific areas and avoidance of nocturnal activity due to concerns about crime. The avoidance of specific areas and nocturnal ventures by women in South Korea is an issue not only of economic significance, as businesses throughout the metropolis must cope with deficient revenue and suboptimal profitability (Fisher 1991;Foster and Giles-Corti 2008), but also an issue of gender equality (Franklin and Franklin 2009;Madriz 1997;Rader and Haynes 2011).
The finding that women tend to be more concerned than men about crime has been of substantial scholarly interest because -with the exception of intimate partner violence, a crime which disproportionately impacts women, with some research indicating that roughly a third of all women on the planet have been the victims of intimate partner violence (Fanslow et al. 2010;Lanier and Maume 2009;Rennison and Rand 2003; World Health Organization 2013)research indicates that victimization rates for violent crime (for example, aggravated assault, homicide) are higher among men than among women. Comparative international research has consistently shown that, regardless of locale, men are more likely than women to be violently victimized (Karmen 2007;LaFree and Hunnicutt 2006). Nonetheless, this study and numerous others conducted in nations around the planet have shown women were more fearful than men of crime.
As to the query of why women tend to be especially fearful of crime, a sizeable body of scholarship suggests gender differences in fear of crime are rooted in patriarchal power structures, and maintained via traditional gender roles and concepts of masculinity and femininity (Meyer and Post 2006;Stanko 1989Stanko , 1990. Within patriarchal societies, being fearful of crime is a component of femininity. To be afraid of crime is to be feminine (Madriz 1997;Stanko 1989). This fear may be affected by a number of socially constructed perceptions of danger -one common explanation for gender differences in fear of crime is that women's fear of crime is enhanced by the thought that any form of personal victimization could escalate into a sexual assault (Ferraro 1996 2013; Pryor and Hughes 2013;Young 1992) -but at the core of such fear are patriarchal power structures (Madriz 1997;Stanko 1989Stanko , 1995. Starting at an early age, women are socialized to behave in traditionally feminine ways (that is, be 'ladylike,' sit with legs crossed, and so on), with one major impetus for such behavior being the avoidance of harm. The basic message inculcated into children is that good things happen to good girls and bad things happen to bad girls. As women mature this message is reinforced in a variety of ways, with rewards (for example, praise, positive attention) for effeminate behaviors, punishments (for example, insults, shunning) for unwomanly behaviors, and blame being assigned to a woman if she is victimized by a man (MacKinnon 1982;Meyers 1997). In her classic text, Nothing Bad Happens to Good Girls, Madriz (1997: 155) argued that fear of crime is an essential component 'in the social control of women because it organizes consent around a strict code of behavior that "good women" need to follow'. Building upon such scholarship, it is reasonable to suggest that the gendered variation in crime-associated trepidation documented herein reflects South Korea's patriarchal power structure.
Historical research indicates that, as early as the first few centuries of the first millenniumwhen the Korean peninsula was governed by the Confederated Kingdoms: Baekje, Silla, and Goguryeo -women in Korea were treated as the property of men. Whereas men were permitted to have multiple wives and/or concubines, women were punished harshly for adultery and even for displays of jealousy. According to Eckert et al. (1990: 21), the 'harsh penalties for female adultery and jealousy are thought to have served the purpose of safeguarding a patriarchal family system characterized by the widespread practice of polygamy'. Centuries later, after the nation was united under the Koryo Dynasty (935-1392) and then the Chosun Dynasty (1392-1910), Korea's patriarchal power structure was fortified as Confucianism became the dominant cultural and official ideology of the nation. Confucian doctrine places an emphasis on the family and social harmony, dictating that (to preserve the harmony of the family) a woman should demonstrate 'Three Obediences': Obedience to the father during childhood, obedience to the husband during adulthood, and obedience to the (eldest) son during widowhood (Cumings 1997;Deuchler 1992;Palley 1990).
In the early 1900s, when Korea was colonized by Japan , the situation and status of women worsened. Korean women were subjugated not only by their relatives and husbands, but also by the Imperial Japanese Army, with somewhere between 100,000 and 200,000 Korean women having been forced into sexual slavery in the 'Comfort Corps'. Korea was liberated in 1945 following the defeat of the Japanese in World War II and the Korean peninsula was divided along the 38th parallel, with Russian forces occupying the north and American forces occupying the south. In 1948, after a few years of governance by the US military, South Korea became a sovereign nation. Influenced by western ideals, South Korea's first constitution included women's suffrage and prohibited discrimination on the basis of gender. The nation was then decimated by the Korean War (1950)(1951)(1952)(1953) but soon thereafter South Korea began rapidly modernizing (Cumings 1997;Eckert et al. 1990;Heo and Roehrig 2014).
Despite the constitutional prohibition of gender discrimination and brisk post-war industrialization, the subordinate status of women persisted. By the late 1980s women still earned, on average, less than half of what men earned. As noted by Palley (1990Palley ( : 1137: 'While the material culture in South Korea has modernized and been affected by Western influences, its behavioral culture maintains and embraces some Confucian traditions … Part of this behavioral culture is reflected in the inequities in women's roles'. Progress toward gender equality in South Korea has been made since the 1980s -of special note, in 2012 Park Geun Hye was the first woman elected to serve as President of South Korea -but gendered power inequities persist. For instance, analyses of gendered income disparities within OECD member nations show that such disparities are higher in South Korea than in any other nation in the OECD. The gender wage gap in South Korea is considerably higher than the gap in Japan, the only other East Asian IJCJ&SD 125 Online version via www.crimejusticejournal.com © 2016 5(4) OECD member nation, and more than twice the gap than in most OECD member nations such as Belgium, France, Italy, Sweden, the United Kingdom and the United States (OECD 2015).
It is plausible to suggest that the gendered differences in fear of crime documented herein reflect the historically entrenched patriarchal power structure in South Korea, with fear of victimization and crime avoidance behaviors serving to limit women's autonomy and maintain patriarchal power relations. Through the avoidance of specific areas, especially at night, or the avoidance of nocturnal activities all together, women may impose restraints on themselves which fuel inequitable power relations between the genders (Madriz 1997;Stanko 1989). As Koskela (1999: 113) summed it up: 'By restricting their mobility because of fear, women unwittingly reproduce masculine domination over space'.
As for policy development, a general implication of this study is that fear reduction strategies should target women. A specific suggestion would be for colleges to offer fear reduction workshops tailored to women college students. This is not to imply that the responsibility for reducing fear of victimization (and/or actual victimization) among women should be placed solely upon women. As pointed out by Stanko (1995), a substantial reduction in fear of crime among women requires a holistic approach to gender equality, inclusive of increased public safety, non-discriminatory employment practices, adequate and affordable child care, and the list goes on. It is simply suggested that workshops designed to reduce women's fear of crime could be implemented as part of the larger struggle toward the evolution of just societies. It is also suggested that the development of fear reduction workshops should be informed by research.
Research indicates that gendered variation in crime-associated trepidation is influenced by socially constructed archetypes, with one common model being the dangerous stranger. Numerous studies have indicated that the tendency for women to be more fearful of crime than are men may be impacted by the belief that criminal victimization is likely, the thought that any criminal victimization could escalate into a sexual assault, and doubts about the ability to handle a criminal perpetrator (Ferraro 1996;Fisher and Sloan 2003;Hickman and Muehlenhard 1997).
To provide a few examples, Lee's (1997: 323) study of fear of crime among South Korean women indicated that the perception that crime was a serious problem was one of the strongest predictors of fear of crime among the women surveyed. Hickman and Muehlenard's (1997: 538) study of fear of rape among college women indicated that, although the probability of acquaintance rape victimization was much greater than the probability of stranger rape, 'when women thought about situations in which they fear being raped, they thought primarily about stranger rape'. Hillinski, Pentecost Neeson, and Andrews' (2011: 119) qualitative study of fear of crime among college women similarly showed that fear of rape was associated with general fear of crime among the women studied, but they also noted that 'some women stated that they were not fearful of crime because they were prepared to defend themselves' (also see Pryor and Hughes 2013). Informed by such research, it is suggested that, to lessen the belief that victimization is likely and increase women's confidence in their ability to repel a perpetrator, fear reduction workshops should include accurate information about the low probability of being violently victimized by a stranger (with an emphasis on the low probability of being sexually assaulted by a stranger) and some basic self-defense training.
Research also indicates there is a reciprocal relationship between fear and constrained behavior (Hickman and Muehlenhard 1997;Liska and Warner 1991). As Liska and colleagues (1988: 835) summed up the relationship, fear 'constrains social behavior and thus results in avoiding such situations, which in turn accentuates the fear'. It is thus suggested that fear reduction workshops also include nocturnal public activities (any activity which encourages the workshop participants to be out in public at night). In brief, providing accurate information on the low likelihood of victimization and basic lessons on self-defense reinforced by nocturnal social outings could enhance the self-confidence of young women to move about the city freely and IJCJ&SD 126 Online version via www.crimejusticejournal.com © 2016 5(4) enjoy themselves rather than remaining shut in their homes confined by fear, with the goal being the initiation of a cyclical process of reduced fear and enhanced autonomy among women. By empowering women to freely move about in public, the workshops may not only lessen fear of crime among women and improve their quality of life, but also reduce what Koskela (1999) treated as the masculine dominance of public space and thus chip away at patriarchal power relations in South Korea.

Research Limitations
While the present findings are of interest, they must be considered with caution owing to methodological issues. First, the surveys were administered to a non-probability sample of college students and it is possible that the youthful composition of the sample shaped the findings. Although there is no consensus in the literature as to the relationship between age and fear of crime -some studies suggest fear increases with age (Toseland 1982), some suggest age has no impact on fear of crime (Eitle and Taylor 2008), and some suggest fear of crime decreases with age (Karakus, McGarrell and Basibuyuk 2010) -it is conceivable that the youth of the respondents affected the present findings. With respect to fear of crime in an East Asian context, Zhang and colleagues' (2009) analyses of data from China showed that younger people were more fearful of crime than were older persons. One particular age-relevant issue to consider is respect for one's elders entrenched in Asian cultures. The respondents' concerns about being victimized while away from home or while out at night may have been imbued by the social vulnerability associated with youth.
Another shortcoming is that the measures of fear and perceived risk were limited, with the measures of fear being offense-specific and the measures of perceived risk being general in nature, distinguishing solely between diurnal and nocturnal perceived risk. While such measures are consistent with the conceptual distinction between 'fear' and 'perceived risk' (Rountree and Land 1996), the questions were limited to a couple of offense-specific fears and did not inquire about multiple offense-specific fears such as fear of aggravated assault, fear of aggravated sexual assault, or fear of aggravated assault with a weapon. Additionally, the only independent variables included in the present analyses were gender and victimization. Based on research which showed concerns about crime may be impacted by variables other than gender and victimization, variables such as age and neighborhood conditions (Eitle and Taylor 2008;Karakus, McGarrell and Basibuyuk 2010), it is probable that concerns about crime in South Korea are affected by variables other than gender and victimization. Finally, there is the fact that fear of crime is a complex phenomenon and quantitative studies, such as the present study, fail to capture the emotional and psychological experiences of fear. Although the bulk of the literature on fear of crime is based on quantitative research, qualitative assessments have shown promise in documenting the causes, nature, and consequences of fear of crime (Hillinski, Pentecost Neeson and Andrews 2011;Meyer and Post 2006;Sur 2014).

Summary remarks
This study is salient for a few reasons. First, this study contributes to the paucity of research on crime-associated concerns in East Asia, inclusive of providing analyses of multiple measures of fear of crime, perceived risk of victimization, and avoidance behaviors. Next, the finding that women were more likely than men to report being fearful, perceiving risk, and engaging in crime avoidance behaviors adds credence to the literature which indicates the relationship between gender and crime-associated anxiety is globally consistent. Finally, the data on crime avoidance behaviors show that, even in a reasonably safe environment such as the Seoul-Gyeonggi Province metropolitan area, concerns about crime may degrade the quality of life enjoyed by women. The finding that roughly half of the women surveyed avoided specific areas and avoided going out at night due to concerns about victimization, despite residing in a metropolitan area with relatively little violent crime, serve as indicators of the strength of patriarchal power relations and the debilitating impact that crime-associated anxiety may have The G-20 is an international convocation of government officials, bankers and business leaders from the world's 20 largest economies. Representatives of the member nations assemble periodically to convene on matters such as the world economy and international economic trade policies and practices. The OECD (Organisation for Economic Cooperation and Development) is a similar assembly of 34 nations, the majority of which of are economically developed nations (for example, Australia, France and the United Kingdom) but a few economically emergent nations (for example, Chile, Mexico and Turkey) are also members. In contrast to the G-20, membership in the OECD is not dependent upon the size of a nation's economy -several nations with sizeable economies (for example, Brazil and China) are not members of the OECD -but rather on the overall development of the economy, with several small advanced nations (for example, the Netherlands and New Zealand) being members of the OECD. 3 The direction of the relationship between victimization and a measure of concern about crime can generally be determined by the positive or negative value of the coefficient, with positive coefficients indicating victimization experience was associated with increased likelihood of having reported a concern about crime. As a precaution, the reader is advised to keep in mind the fact that logistic regression models are inherently interactive. In the present study, this fact is important when examining the coefficients for the physical assault measure. In preliminary regression analyses of the data (results not shown) the tendency for assault victimization to be associated with low fear/perceived risk was consistent but, as additional variables were included in the regression models, the values of the coefficients changed (as is inevitable with logistic regression). The coefficients for the physical assault victimization variable changed considerably, going from negative to positive. As the impact of assault victimization on the various measures of concerns about crime was not statistically significant, this issue of no relevance and has been noted solely to caution the reader against becoming distracted by the positive values of the assault victimization coefficients in Table 4.