WHY DO PEOPLE COMPLY AND COOPERATE WITH THE POLICE? A CULTURAL EXPLANATION By Sung Uook Lee A DISSERTATION Submitted to Michigan State University in partial fulfillment of the requirements for the degree of Criminal Justice—Doctor of Philosophy 2020 WHY DO PEOPLE COMPLY AND COOPERATE WITH THE POLICE? A CULTURAL ABSTRACT EXPLANATION By Sung Uook Lee The current dissertation examines police legitimacy in the South Korean context. Prior research has tended to focus on the relational aspects of police-citizen relationships and is generally suggestive of an important role for a variety of relational constructs. Although important, this work has tended to pay less attention to person-level constructs within individuals. Furthermore, prior police legitimacy research lacks examination of cultural values as well and only focuses on police-citizen encounters. Additionally, although recent studies regarding South Korean public perception of police legitimacy have been emerging, more extensive investigation is needed. The primary goal of the current study is to examine the impact of propensity to trust and Confucian values on perceived police legitimacy, operationalized here using the Integrated Framework of Legitimacy (Hamm et al, 2017). To this end, the current dissertation uses data collected from South Korean university students to contribute to the literature (1) an evaluation of the role of person-level constructs in predicting public perceptions of police legitimacy and (2) a first test of the IFL in the South Korean context. ACKNOWLEDGEMENTS I would like to express my sincere gratitude to everyone who supported me during my academic development at Michigan State University. I would like to thank Dr. Joseph Hamm, my advisor and chair of my dissertation committee, for all the guidance and encouragement for me to push through my academic struggles. His openness and passion for guidance have been an inspiration. By setting him as my role model, I hope to achieve his level of academic and pedagogical success. I would also like to thank Dr. Mahesh Nalla, the advisor and co-chair, for steering me to the right academic path. Without his wise guidance, I would have been lost. I would like to also thank the committee members, Dr. Charles Corley and Dr. Steven Gold for their support. My special thanks go to my family who has supported and waited through my academic development. To my mother who has sacrificed her life for my success, thank you for being there for me to rely on. To my father who has been my role model both as an honorable man and as a professor, thank you for being there for me to look up to. iii TABLE OF CONTENTS LIST OF TABLES ......................................................................................................................... vi LIST OF FIGURES ...................................................................................................................... vii CHAPTER 1. INTRODUCTION ....................................................................................................1 Statement of the Problem ...........................................................................................................1 The Current Dissertation ............................................................................................................4 CHAPTER 2. POLICE LEGITIMACY .........................................................................................6 Police Legitimacy .......................................................................................................................6 Sources of Legitimacy ...........................................................................................................6 Integrated Framework of Legitimacy .........................................................................................7 Stage 1 ....................................................................................................................................8 Procedural Justice .............................................................................................................8 Distributive Justice............................................................................................................9 Police effectiveness ...........................................................................................................9 Stage 2 ..................................................................................................................................10 Trustworthiness ...............................................................................................................10 Normative Alignment .....................................................................................................11 Stage 3 ...................................................................................................................................11 Trust ................................................................................................................................11 Obligation to Obey ..........................................................................................................12 Stage 4 ...................................................................................................................................12 Compliance .....................................................................................................................13 Cooperation ....................................................................................................................13 CHAPTER 3. PERSON-LEVEL CONSTRUCTS ........................................................................15 Person-Level Constructs ...........................................................................................................15 Propensity to Trust ...............................................................................................................15 Confucian Values .................................................................................................................16 Social Hierarchy..............................................................................................................16 Social Harmony ..............................................................................................................17 Face-Saving.....................................................................................................................17 Humbleness .....................................................................................................................18 Integrating the Person Into IFL .................................................................................................18 Stage 1 ..................................................................................................................................19 Stage 2 ..................................................................................................................................19 Stage 3 ..................................................................................................................................20 Stage 4 ..................................................................................................................................21 iv CHAPTER 4. PRESENT STUDY .................................................................................................22 Methods.....................................................................................................................................22 Research Cite .......................................................................................................................23 Data Collection ....................................................................................................................23 Survey .............................................................................................................................25 Analytic Strategy .................................................................................................................29 CHAPTER 5. RESULTS ...............................................................................................................32 Demographics ...........................................................................................................................32 Univariate Analyses ..................................................................................................................33 Mean Differences .....................................................................................................................35 Confirmatory Factor Analyses ..................................................................................................37 Single Factor CFAs: IFL Constructs....................................................................................38 Single Factor CFAs: Person-Level Constructs ....................................................................41 Multifactor CFA...................................................................................................................43 Integrated Framework of Legitimacy ..............................................................................43 Person-Level Constructs ..................................................................................................45 All Constructs ..................................................................................................................46 Structural Regression ................................................................................................................49 Integrated Framework of Legitimacy ..................................................................................49 Propensity to Trust IFL ...................................................................................................50 IFL .......................................................................................................53 Social Hierarchy Social Harmony IFL........................................................................................................55 IFL ..............................................................................................................57 Face-Saving Humbleness IFL ..............................................................................................................59 CHAPTER 6. DISCUSSION .........................................................................................................62 Key Findings ............................................................................................................................62 Propensity to Trust ..............................................................................................................62 Social Hierarchy ..................................................................................................................65 Social Harmony ..................................................................................................................68 Face-Saving .........................................................................................................................69 Humbleness .........................................................................................................................71 Limitations ...............................................................................................................................72 6.3 Conclusion .........................................................................................................................75 APPENDICES ..............................................................................................................................77 APPENDIX A. Survey Instrument English .............................................................................78 APPENDXI B. Survey Instrument Korean ...............................................................................86 REFERENCES .............................................................................................................................97 v LIST OF TABLES Table 1. Description of Universities ..............................................................................................24 Table 2. Demographics ..................................................................................................................32 Table 3. Description of Study Variables ........................................................................................34 Table 4. Mean Differences ............................................................................................................36 Table 5. CFA of IFL ......................................................................................................................43 Table 6. IFL Latent Constructs Correlation Matrix .......................................................................44 Table 7. CFA Person-Level Constructs .........................................................................................45 Table 8. Person-Level Constructs Correlation Matrix ...................................................................46 Table 9. CFA All Latent Constructs .............................................................................................46 Table 10. All Latent Constructs Correlation Matrix ......................................................................48 Table 11. Structural Regression IFL ..............................................................................................49 Table 12. Structural Regression: Propensity to Trust IFL ........................................................52 Table 13. Structural Regression: Social Hierarchy IFL ............................................................54 Table 14. Structural Regression: Social Harmony IFL .............................................................56 Table 15. Structural Regression: Face-Saving IFL ...................................................................58 Table 16. Structural Regression: Humbleness IFL ...................................................................60 vi LIST OF FIGURES Figure 1. Stages of IFL ....................................................................................................................7 Figure 2. Distribution of Provinces and Metropolises in South Korea ..........................................25 Figure 3. Structural Regression Paths: Propensity to Trust IFL ..............................................65 Figure 4. Structural Regression Paths: Social Hierarchy IFL ..................................................68 Figure 5. Structural Regression Paths: Social Harmony IFL ...................................................69 Figure 6. Structural Regression Paths: Face-Saving IFL .........................................................70 Figure 7. Structural Regression Paths: Humbleness IFL .........................................................72 vii Statement of the Problem CHAPTER 1. INTRODUCTION Although the police can obtain compliance and cooperation from the public via coercive methods, it has been argued that voluntary compliance and cooperation works better (Weber, 1968). In particular, when individuals view the police to be a legitimate (that is, rightful) authority, it will lead to voluntary compliance and cooperation (Sunshine & Tyler, 2003). Because of legitimacy’s importance in producing self-regulating and law-abiding behavior, research has extensively investigated the construct (Hinds, 2007; Hinds & Murphy, 2007; Levi, Sacks & Tyler, 2009; Reisig, Bratton & Gertz, 2007; Reisig & Lloyd, 2009; Sunshine & Tyler, 2003). Prior research describes legitimacy in similar ways. Beetham (1991) suggests that legitimacy is the willingness of the public to defer to authority and the law via voluntary compliance and cooperation. Tyler and Huo describe legitimacy as the belief that a legal authority is entitled to be obeyed and deferred to. In other words, legitimacy is generally understood to be a voluntary willingness to defer to an authority because they are believed to be a rightful authority. When people believe that the police are legitimate, they are more likely to voluntarily comply with police orders and cooperate with the police (Blader & Tyler, 2003; Baker & Gau, 2018; Gau, 2015; Murphy, Bradford & Jackson, 2016; Nagin & Telep, 2017; Trinkner & Tyler, 2018; Woo, Maguire & Gau, 2018). The police agencies in the US have historically tended to emphasize coercive compliance. Deterrence Theory argues that if the cost outweighs the benefit, individuals will not engage in criminal behavior (see Klepper & Nagin, 1989; Nagin & Paternoster, 1991; Nagin & Pogarsky, 2003; Paternoster, 1987, 1989; Paternoster & Iovanni, 1986; Paternoster, Saltzman, Chiricos & Waldo, 1982; Paternoster, Saltzman, Waldo & Chiricos, 1983; Sherman, 1993). 1 Individuals, therefore, comply with the law and the police because they fear sanctions, not because of the institution’s legitimacy. However, prior research on deterrence is inconclusive regarding its effectiveness with some studies finding that increases in the perceived certainty of punishment have no impact on reducing criminal behavior (Paternoster, 1987; Tyler, 1990). Even when deterrence effects have been found, they are often relatively small (MacCoun, 1993; Nagin, 1988). As such, Deterrence Theory has challenges in promoting the public’s compliance and cooperation. Tyler offers a different perspective in promoting compliance and cooperation that focuses on fair treatment during police-citizen interaction. Tyler’s Process-Based Theory of Legitimacy (1990; 2001; 2006) suggests that fair treatment (i.e., procedural fairness) by the police is the primary driver of increases in public perceptions of legitimacy. Scholars have extensively researched the Process-Based Theory of Legitimacy and added valuable knowledge to police legitimacy literature. Significant work, then, has investigated legitimacy, procedural fairness, and a variety of other related constructs. Although important, this work is typically inconsistent in how it organizes these constructs leading to important debates about the nature of police legitimacy (see Bottoms & Tankebe, 2012; Tankebe, 2013; Tyler & Jackson, 2013). To aid in organizing the police legitimacy scholarship, Hamm and colleagues (2017) developed an Integrated Framework of Legitimacy (IFL) that adopts and builds upon Tyler’s Process-Based Theory of Legitimacy. The IFL argues that, in place of a singular construct, police legitimacy can be understood as a process of four stages. Stage 1 proposes that the conceptual starting point of legitimacy is an evaluation in which people assess the information from direct or vicarious interactions with the police. In Stage 2, people use these evaluations to assess the police themselves. In Stage 3, these 2 evaluations drive internalizations whereby the individual starts to feel or believe things about their relationship with the police (e.g., I am afraid of them, I trust them, I should obey them). In Stage 4, these individual internalizations drive reactions like (non)compliance and (non)cooperation. To date, most of the literature has focused on the relational aspects of police-citizen interaction (e.g., Reisig, Bratton & Gertz, 2007; Reisig & Lloyd, 2009; Sunshine & Tyler, 2003). Although examining the relational aspects are important, not much attention has been focused on the person-level constructs. Pickett and colleagues (2018) highlight this oversight in their work on relational schemas. They argue that these schemas foster interpersonal behavioral tendencies or dispositions regarding how one should behave Person-level constructs may be vital in assessing police legitimacy. Person-level constructs are predisposed behavioral tendencies which drive behavior in ways that do not depend on the situation. Propensity to trust is one such construct that addresses generalized expectations about the trustworthiness of others (Rotter, 1971). In particular, propensity to trust is a predisposed or baseline tendency that makes individuals more willing to trust others, regardless of the specific characteristics or behavior of the individual being trusted (Mayer et al., 1995; Frazier et al., 2013). Individuals with higher propensity to trust are more likely to report trust in all target and therefore should feel more trust toward the police, and ultimately, report higher compliance and cooperative behavior for reasons beyond the relational concerns that have traditionally been the focus of legitimacy research. Confucianism is another person-level construct that heavily influences many Asian societies. This collection of cultural values suggests that there are certain ways an individual should act in society, such as respecting elders, maintaining order, conforming to the social norm, and respecting authority (Tan & Chee, 2005; Chao, 1994). The central themes of 3 Confucianism are, hierarchical interpersonal relationships, conforming to social norms, compliance, maintenance of social order and harmony (Park & Chesla, 2007; Tan & Chee, 2005). Fundamentally, Confucianism argues that individuals should comply and cooperate with the authority which, because of their place as a salient social authority, is likely to be relevant to police legitimacy and especially felt obligations to cooperate and comply. The Current Dissertation This dissertation seeks to contribute to the legitimacy literature by examining the role of person-level constructs within the South Korean context. Although Tylerian approaches to police legitimacy have generally been supported in research conducted in Western countries such as USA (Gau, 2011, 2013, 2014; Gau et al., 2012; Reisig et al., 2007; Sunshine & Tyler, 2003; Tyler & Huo, 2002; Tyler & Wakslak, 2004; Wolfe et al., 2016), Australia (Mazerolle et al., 2013; Murphy, 2005; Murphy et al., 2014), Canada (Penner et al., 2014), research in non- Western settings such as Ghana (Tankebe, 2009), Jamaica (Reisig & Lloyd, 2009), Japan (Tsushima & Hamai, 2015), China (Sun et al., 2017), Hong Kong (Cheng, 2015), and Nigeria (Akinlabi, 2017), have yielded only mixed support. This study follows the lead of the IFL in conceptualizing legitimacy which, because of its strong footing in Tylerian approaches, focuses primarily on the interactions between law enforcement and the public. Although important, this neglects the potential role of person-level constructs. This dissertation will therefore assess the impact of a person-level construct that has been evaluated in a variety of contexts (propensity to trust) and a specific cultural concern in Asian societies (Confucian values) on police legitimacy in the South Korean context. To this end, the current dissertation hypothesizes that propensity to trust (Hypothesis 1) and the specific Confucian values of social hierarchy (Hypothesis 2), social 4 harmony (Hypothesis 3), face-saving (Hypothesis 4), and humbleness (Hypothesis 5) will add to the predictive utility of the IFL. The following chapters review police legitimacy (Chapter 2) and the potential role of person-level constructs (Chapter 3). Chapter 4 then presents the current study’s methodology and Chapter 5 reports the results of this study. The dissertation concludes with Chapter 6 which highlights the important findings of this research as contextualized against their theoretical and practical implications. 5 Police Legitimacy CHAPTER 2. POLICE LEGITIMACY Legitimacy has long been understood to be important in a variety of governance contexts. Weber (1968) argues that effective and efficient governance requires voluntary compliance from the public and that legitimacy is a major element of this. Scholars have argued that obedience does not best come from a coercive authority that bases its power on sanctions; rather, this power should come from voluntary compliance rooted in people’s perception of the legitimate authority of the government (Aberbach & Walker, 1970; Hanson & Fowler, 1971; Paige, 1971; Wright, 1976). Considerable research has examined the relationship between legitimacy, compliance, and cooperation in the policing context and suggests that increases in the public’s perception of legitimacy lead to increases in compliant and cooperative behavior (Hough et al., 2010; Murphy et al., 2009; Murphy et al., 2016; Reisig et al., 2014; Sunshine & Tyler, 2003; Woo et al., 2018) and decreases in offending behavior (Jackson et al., 2012; Dai, Frank & Sun, 2011; Kochel, 2012; Murphy & Cherney, 2012; Reisig et al., 2007). Sources of Legitimacy. One of the major foci within the police legitimacy literature is on understanding the sources of legitimacy. The Process-Based Theory of Police Legitimacy postulates that people are more likely to accept the decisions made by the police when they feel that the police used fair procedures (Tyler & Huo, 2002; Sunshine & Tyler, 2003). Prior research has identified procedural fairness as the main driver of legitimacy (Cremer & Tyler, 2007; Gau, 2014; Hinds, 2007; Sunshine & Tyler, 2003; Tyler & Wakslak, 2004). Mazerolle and colleagues (2013) found that experienced procedural justice not only was the primary driver of perception of police legitimacy in the specific experience but also a primary driver of the perceived legitimacy of the entire police force. Scholars have also found significant support for a variety of other 6 constructs like distributive justice (Antrobus et al., 2015; Tyler & Fagan, 2008; Tyler & Huo, 2002; Sunshine & Tyler, 2003), police effectiveness, (Bottoms & Tankebe, 2012; Tankebe, 2013; Hough et al., 2010; Jackson et al., 2012; Jonathan-Zamir & Harpaz, 2014; Murphy et al., 2015; Rosenbaum et al., 2015; Sargeant et al., 2014; Wolfe et al., 2016), and trustworthiness as a source of legitimacy (DeCremer & Tyler, 2007; Nagin & Telep, 2017; Reisig et al., 2012; Sunshine & Tyler, 2003). Although important, this work has tended to struggle in consistently conceptualizing these constructs and, especially, the expected relations among them. (Gau, 2011; Tyler, 2004). Different studies have organized the constructs of legitimacy differently. To this state of affairs, Hamm and colleagues (2017) present the Integrated Framework of Legitimacy which postulates that legitimacy can be best understood as a process and not as a single construct. Through this lens, the IFL sought to organize the many constructs of police legitimacy into four stages and present theoretical arguments regarding the relations among them. Integrated Framework of Legitimacy The central idea of the Integrated Framework of Legitimacy (IFL) is that, in place of a singular construct, legitimacy can be understood as a process comprised of four linked stages such that evaluations of the interaction with the police such as procedural fairness and distributive fairness (Stage 1) drive the evaluations of the police via constructs like the trustworthiness (Stage 2) that, in turn, drive internalizations within the individual such as trust (Stage 3) that lead to reactions such as cooperation and compliance (Stage 4; see Figure 1). Figure 1. Stages of IFL Stage 1. Evaluation of the Interaction • Procedural Justice • Effectiveness • Distributive Justice Stage 2. Evaluation of the Target • Trustworthiness • Normative Alignment Stage 3. Internalization Stage 4. Reaction • Trust • Obligation to Obey • Cooperation • Compliance 7 Stage 1. In the first stage of the IFL, evaluation of the interaction, individuals gather information from interactions with law enforcement and evaluate them. In this information- gathering stage, they gather information by asking questions such as: was the interaction with the police procedurally fair? Did the police distribute justice fairly? Are the police effective in controlling crime and maintaining order? Three constructs that fall into this stage of the IFL are procedural justice, distributive justice, and police effectiveness. Procedural Justice. Thibaut and Walker (1975) examined how litigants’ satisfaction with the decision was influenced by the perception of the fairness of the dispute resolution process. Their research found that, regardless of the outcome of the dispute resolution, when the litigants perceived that the processes of the dispute resolution were fair, they were more likely to report being satisfied (Thibaut & Walker, 1975). Tyler adopted this idea of procedural justice and applied it to criminal justice settings and law enforcement in particular. He argued that individuals are more likely to accept the decisions made by the authority when they perceive that the process by which that decision is reached was fair (Tyler, 1990; Tyler & Fagan, 2008; Sunshine & Tyler, 2003). Prior research supports this the link between procedural justice and police legitimacy (Baker & Gau, 2018; Barworth & Murphy, 2015; Jackson et al., 2013; Jackson et al., 2018; Madon et al., 2017; Murphy et al., 2013; Reisig & Lloyd, 2009; Tankebe et al., 2016). Gau (2015) examined procedural justice, police legitimacy, and legal cynicism and found that procedural justice not only enhances people’s perceptions of police legitimacy but also elevates their opinions about the law in general. Prior research on procedural justice on non- American samples had similar findings as American samples (Reisig et al., 2014; Sargeant et al., 2014; Woo et al., 2018). Antrobus and colleagues (2015) found that procedurally just policing 8 appears to have a greater effect on the public’s sense of obligation to obey the police and the perception of legitimacy. Distributive Justice. Distributive justice focuses on the fairness of outcomes (Sarat, 1977). When the public views the police officers to be handing down justice equally across all groups of people, the public views the police to be a legitimate authority (Tankebe et al., 2016; Sunshine & Tyler, 2003). Prior research also supports the association between distributive justice and police legitimacy (Hinds, 2009; Reisig et al., 2007; Sunshine & Tyler, 2003; Tyler & Fagan, 2008). Tyler and Huo (2002) found that American people are more accepting of the decisions made by the police when they felt that the outcomes of those decisions are fairly distributed. In the South Korean context, Woo and colleagues (2018) also found that distributive justice was a significant predictor of compliance. In the Australian context, distributive justice was strongly associated with the perception of legitimacy (Hinds & Murphy, 2007; Murphy, Hinds & Fleming, 2008). Police Effectiveness. When the public views the police to be effective in controlling crime and maintaining social order, people are more likely to view the police as legitimate (Tyler, 2005; Wilson & Kelling, 1982). Sunshine and Tyler (2003) argue that “police develop and maintain legitimacy through their effectiveness in fighting crime and disorder in the community” (p. 514). Tankebe and colleagues (2016) found that police legitimacy is grounded in the judgments of police effectiveness. Studies have identified that the public perception of police effectiveness does impact the perception of police legitimacy. Levi and colleagues (2009) found that people who viewed the courts or the police to be competent in their job were more likely to view the police to be a legitimate authority. Wolfe and colleagues (2016) found that police effectiveness had a significant positive relationship with the perception of police legitimacy. 9 Studies in different cultural contexts presented similar findings. In the Slovenian context, Reisig and colleagues (2014) found that people viewed police to be a legitimate authority and had higher moral credibility when they perceived them to be effective in their job. In the Australian context, Sargeant and colleagues (2014) found that police effectiveness was a significant predictor of police legitimacy for Australians of Vietnamese and Indian descent. In the South Korean context, Woo and colleagues (2018) reported that police effectiveness was a significant predictor of people’s willingness to cooperate. Stage 2. The second stage of the IFL is the evaluation stage. The IFL postulates that the information gathered in Stage 1 informs our evaluation of the police (Hamm et al., 2017). Two constructs in this second stage age trustworthiness and normative alignment. Based on the information gathered in Stage 1 such as procedural justice, distributive justice, and police effectiveness, individuals determine the trustworthiness and normative alignment of the police. Thus, to evaluate the police based on the information gathered in Stage 1, they ask questions such as, does the character of the police in my community make them worthy of trust? Do police share similar values, norms, and goals with the community? Trustworthiness. Trustworthiness concerns the character of the target, in this case the character of the police. Trustworthiness is defined by PytlikZillig and colleagues (2016) as “beliefs, evaluations, or expectancies of the target that are often theorized to form the basis for trust” (p. 10). Mayer and colleagues (1995) conceptualized trustworthiness as having three major dimensions: 1) ability, 2) integrity, and 3) benevolence. According to Colquitt and colleagues (2007), ability addresses whether police officers have the skills needed to do their job; integrity focuses on whether police officers have the moral and ethical principles to do their job; benevolence addresses whether a police officer is believed to want to do good for the 10 community. If an individual believes the police to be trustworthy, they are more likely to follow the law, accept the decisions made by the police and are more likely to cooperate with the police (Tyler & Huo, 2002). Studies have examined the connection between trustworthiness and legitimacy (Gau, 2014; Nagin & Telep, 2017; Reisig et al., 2007; Sunshine & Tyler, 2003). Levi and colleagues (2009) found that trustworthiness was one of the main drivers of perception of police legitimacy. Normative Alignment. According to Huq and colleagues (2016), normative alignment refers to citizens’ belief that the police generally share societal expectations with their communities. People who feel normatively aligned with the police believe that their power is justified which leads to the perception of legitimacy (Jackson et al., 2012). Studies have identified normative alignment as one of the drivers of legitimacy (Huq et al., 2016; Jackson et al., 2012; Trinkner & Tyler, 2018; Tyler & Jackson, 2014). Gerber and Jackson (2016) suggest that when individuals feel normatively aligned with the police, they believe that the officer will use their authority that is in accordance with social expectations or socially desirable and adequate conduct. Stage 3. In Stage 3 individuals internalize their evaluation of the police. Although not exclusive, two constructs in this stage are trust and obligation to obey. In this stage, individuals internalize a sense of obligation and a willingness to accept vulnerability toward the police based on the evaluations formed in Stage 2. Thus, if they believe the police to be trustworthy and normatively aligned, they decide to trust the police and feel a sense of obligation to obey the police because of our positive evaluation of them in Stage 2. Trust. Trust is defined as “a psychological state of the trustor characterized by a willingness to rely upon, give control to, support, or otherwise be vulnerable to the trustee” 11 (PytlikZillig et al., 2016, p. 10; see also Mayer et al., 1995). The main point of the trust is the willingness to take a risk. Therefore, the level of trust impacts the amount of risk a person is willing to take in the trusting relationship (Mayer et al., 1995). In the policing context, when individuals trust the police, they are willing to accept vulnerability because they trust the police to be a legitimate authority. Prior studies have identified trust as one of the drivers of legitimacy (Hamm et al., 2017; Jackson & Gau, 2016). Hough and colleagues (2010) found that trust in the police is a significant predictor of perceived legitimacy and also suggest that the relationship between trust and legitimacy is stable across different cultures. Obligation to Obey. Another important aspect of police legitimacy is the fact that people feel an obligation to obey the law and the police. People comply with the orders of the police and the law because they believe it is the right thing to do (Jackson et al., 2012; Tyler, 2006; Tyler & Huo, 2002). However, this normative belief of the “right thing to do” does not occur if the public does not view the police to have the right to govern their behavior (Beetham, 1991; Tyler, 2006; Bottoms & Tankebe, 2012). If individuals evaluate interactions with the police themselves as a rightful authority, they will feel that it is the right thing to obey the police (Gau & Brunson, 2010). Prior studies have identified connections between the obligation to obey and police legitimacy (Antrobus et al., 2015; Gau, 2014; Jackson et al., 2012; Reisig et al., 2012; Wolfe et al., 2016; Woo et al., 2018). Gau (2015) found that legitimacy was a positive predictor of obligation to obey. Stage 4. In Stage 4, individuals react to the internalizations from Stage 3. There may be different types of reactions, but two particular types of reactions, compliance and cooperation are examined in the IFL. After individuals internalize their feelings in Stage 3, they react to those feelings by complying and cooperating. Thus, if individuals have positive internalized feelings 12 such as trust towards the police and a sense of obligation to obey the police, this internalized feeling will lead to compliance and cooperation with the police. Compliance. Although the police can obtain compliance and cooperation from the public via coercive methods, it has been argued that voluntary compliance and cooperation works better (Weber, 1968). In particular, when individuals view the police to be a legitimate (that is, rightful) authority, it will lead to voluntary compliance (Sunshine & Tyler, 2003). People comply because they believe the law and the police are just and it is the right way to behave based on their social norm (Tyler, 2006). Individuals with positive perception of police legitimacy are more likely to comply with the police via law-abiding behaviors such as not breaking traffic laws, legally disposing trash, parking legally, etc. (Sunshine & Tyler, 2003). Studies have found a strong association between compliance and police legitimacy (Barworth & Murphy, 2015; Fagan & Tyler, 2005; Hinds, 2009; Hough et al., 2010; Jackson et al., 2012; Murphy et al., 2008). Specifically, findings suggest that procedural justice and distributive justice promote compliant behavior with the police (Levi et al., 2009; Reisig et al., 2007; Sunshine & Tyler, 2003). Cooperation. Sampson, Raudenbush and Earls’ (1997) claim that to effectively control crime and keep order, the police depend on public cooperation. Individuals who view the police as legitimate are more willing to cooperate with the police by reporting crimes or identifying criminals (Mazerolle et al., 2014; Murphy et al., 2015; Sunshine & Tyler, 2003). Individuals with higher levels of perception of police legitimacy are more likely to cooperate with the police via reporting crimes and providing information about a crime or a suspect. Studies have found a strong association between cooperation and police legitimacy (Hough et al., 2010; Reisig et al., 2007; Reisig & Lloyd, 2009; Rosenbaum et al., 2015; Woo et al., 2018). People who view the police as legitimate are more willing to cooperate with the police by reporting crimes and 13 helping to identify suspected criminals (Sunshine & Tyler, 2003). Furthermore, positive perception of procedural justice, police effectiveness, and trust in police lead to higher levels of cooperative behavior (De Cremer & Tyler, 2007; Murphy et al., 2015). 14 CHAPTER 3. PERSON-LEVEL CONSTRUCTS Previous research addressing the Process-Based Theory of Police Legitimacy research has tended to focuses on the relational aspects of police-citizen encounters (e.g., Gau & Brunson, 2010; Hough et al., 2010; Kochel et al., 2013; Murphy & Tyler, 2008; Murphy et al., 2015; Reisig et al., 2007). There has been a lack of attention in this work on person-level constructs which may be important because they are predispose attitudes and behaviors in ways that do not depend on the specific situation. When specific information is not available, they may entirely base our thoughts about the police on these predispositions but even when specific information is available, these constructs likely color them and our behavioral reactions. Also, constructs of police legitimacy, may be culturally sensitive. Tan and Chee (2005) argue that different cultures may have different responses to authority. For example, the development of trust in subordinates is influenced by values such as deference to authority, in part due to the Confucian emphasis on the social hierarchy. Person-Level Constructs Propensity to Trust. One of person-level constructs addressed here is propensity to trust (also sometimes called dispositional trust). Mayer and colleagues (1995) argue that individuals have a predisposition that makes them more or less willing to trust others generally. Thus, individuals with high trust propensity are more likely to trust regardless of a particular situation (Mayer et al., 1995). This person-level construct is important because propensity to trust will influence how much a person generally trusts a trustee prior to obtaining information regarding the trustworthiness of the trustee (Hamm et al., 2017; Mayer et al., 1995) which is especially important when individuals lack knowledge or experience with the police (Hamm et al., 2013). Propensity to trust varies, not only within but also across cultures (Gefen & Heart, 2006; 15 Greenberg et al., 2008; Tan & Sutherland, 2004). Jarvenpaa and Tractinsky (2003) suggest that individuals from individualistic cultures are more likely to trust others until they have a reason not to trust while individuals from collectivistic cultures are more likely to trust based on relationships and first-had knowledge/experiences. Confucian Values. Originating in China, Confucianism has infused itself into the fabric of many Asian societies, including South Korea, Japan, and China (Chao, 1994). Confucianism is a body of beliefs, rituals, and codes of conduct taught by the Chinese philosopher Confucius around in 500 BC (Tan & Chee, 2005). Central themes of Confucianism are a strong emphasis on harmony and the hierarchical arrangement of interpersonal relationships (Park & Chesla, 2007; Tan & Chee, 2005). According to Confucian ideology, stability in society is based on ordered relationships (Tan & Chee, 2005). Thus, there are certain ways individuals should act and behave in relation to others which are rooted in social hierarchies that are of utmost importance in ensuring social cohesion. In South Korea, Confucian values emphasize the importance of family, social hierarchy, social harmony, conforming to norms, hard work, and education (Park & Shin, 2006; Yim, 2002). Confucian values generally emphasize respect for elders, parents, ancestors; family loyalty; respect and submission towards authority based on seniority; the importance of the group over the individual; and the importance of social order over personal freedom (Levy, 1992; Park & Shin, 2006; Shin & Shaw, 2003). Although there are more than four dimensions of Confucianism, the current study focuses on the four dimensions discussed below. Social Hierarchy. The first dimension of Confucian values is social hierarchy, or deference to authority. As noted above, Confucian traditions argue that all human relations are hierarchical. A young person should respect and obey their elder, a son/daughter should obey 16 their parents, a student should obey their teacher, and a civilian should obey the government and its agents (Na & Cha, 1999; Park & Shin, 2006; Tan & Chee, 2005). Deference to authority within hierarchical interpersonal relations is a core facet of Confucian values. It is, therefore, is the general norm to trust, obey, comply, and cooperate with any authority, including police officers. Dalton and Ong (2005) found that East Asian countries were much more likely than Western countries to respect authority, obey teachers, and follow instructions, respect parents and elders. Zhai (2017) concluded that both Japanese and Chinese people who emphasized Confucianism were more likely to blindly defer to the political authority and accept social hierarchies. Social Harmony. The second dimension of Confucian values is social harmony. In the Confucian tradition, avoiding conflict is especially important. Behaviors that cause interpersonal conflict such as excessive self-assertion are considered to be shameful and are regarded as disruptions to social harmony (Park & Shin, 2006). Confucian traditions view yielding and conceding for the sake of interpersonal harmony as good behavior (Shin & Shaw, 2003). The goal of social harmony is to avoid conflict among members of society. Because of the main emphasis on maintenance of the social harmony of Confucian values, individuals are pressured to not disrupt the social harmony and maintain order (Chung et al., 2008). Furthermore, Chung and colleagues (2008) suggest that individuals with Confucian values such as social harmony place much more emphasis on universal ethical and moral codes. Individuals with more social harmony beliefs are more likely to abide by the law and follow the universal ethical and moral code of the society by not disrupting it. Face-Saving. The third dimension of Confucian values is face-saving. Individuals in collectivistic cultures put great emphasis on ideologies such as face-saving (Ting-Toomy et al., 17 1991; Yau, 1988). Keeping or saving one’s face is especially important in the Confucian society because when one’s face is lost, then one’s moral integrity and their family’s integrity is also lost (Chung & Pysarchik, 2000). Face-saving refers to maintaining an individual’s dignity and avoiding threats to one’s public image (Monkhouse, Barnes, & Pham, 2013). As a result, members of the Confucian culture are always under strong pressure to act in accordance with the social norm to save face (Chung & Pysarchik, 2000). To save or maintain one’s reputation, individuals in Confucian society will act according to the social norm and trust, obey, comply, and cooperate with the police because failing brings dishonor. Humbleness. The fourth dimension of Confucian values is humbleness. In Confucian teachings, a person should keep a low profile to maintain harmony in the community (Tu, 1998). Rowatt and colleagues (2006) suggest that individuals with humbleness show behaviors such as respectfulness, willingness to admit imperfections, lack of self-focus, and modesty in social behavior. Humbleness is also closely associated with higher cooperative behavior and lower delinquent tendencies (Hilbig & Zettler, 2009; Hilbig et al., 2013; Zettler, 2013). Ting-Toomy and colleagues (1991) find that more humble individuals are more likely to engage in obliging and conforming behaviors. As such, individuals with more humbleness may be more likely to oblige to others and conform their behavior toward the law and the police. Integrating the Person into IFL Research addressing propensity to trust and Confucian values suggests that people may have different reactions to and thoughts about the police as a function of individual differences on these variables. However, these person-level constructs have not been tested in police legitimacy research yet. In particular, research addressing the IFL has failed to fully examine 18 how person-level constructs may impact police legitimacy. The current study seeks to address this gap. Stage 1. As noted above, Stage 1 starts with the gathering of information. In this stage, individuals evaluate information from their personal and vicarious interactions with the police. This is likely to be impacted by person-level constructs such as propensity to trust and Confucian values. The idea is that invidual differences may influence how people view the same behavior from the police differently. One simple explanation could be that, individuals with higher levels of propensity to trust have a higher tendency to trust in general (Mayer & Davis, 1995). One may simply view the police interaction to be more positive because they have higher propensity to trust. Individuals with different levels of Confucian values may also gather information differently. Deference to authority within hierarchical interpersonal relations is a core facet of Confucian values (Chung et al., 2008). An individual with a higher social hierarchy tendency may be more willing to view the police legitimately which may positively influence the interaction with the police. Individuals with higher social harmony, face-saving, and humbleness tendencies may be more inclined to view the interaction with the police positively because they do not disrupt the harmony. Stage 2. In Stage 2, individuals evaluate the police based on the information they have gathered in Stage 1. This evaluation is likely to be impacted by person-level constructs such as propensity to trust and Confucian values. Theilmann and Hilbig (2014) suggest that individual differences impact trustworthiness. Specifically, individual differences in honesty and humility have resulted in significantly different trustworthiness tendencies such that people with higher honesty and humility tend to act in ways consistent with greater trustworthiness. These individuals, in turn, tend to expect others to be trustworthy (Berg et al., 1995; Krueger et al., 19 2008). If one has different levels of propensity to trust and Confucian values, an individual may have different levels of trustworthiness tendencies. Individuals with a higher propensity to trust assume others are generally trustworthy as well (Alarcon et al., 2016; Butler, 1999). Contrarily, individuals with low propensity to trust are cynical and skeptical with low levels of trustworthiness (Chatman, 1991). Confucian values put a strong emphasis on social relations. Individuals with strong social hierarchy and strong social harmony tendencies are more likely to value social relations in the Confucian society which leads to higher trustworthiness tendency (Tan & Chee, 2005). Stage 3. Individuals internalize the evaluations they have made in Stage 2. This internalization of trust and obligation to obey feelings is likely to impacted by person-level constructs such as propensity to trust and Confucian values. Studies suggest that propensity to trust influences trust (psychological trust) because a person with a high propensity to trust assumes that most people are fair, honest and have good intentions and therefore they are more likely to trust others (Costa & McCrae, 1992; Mooradin et al., 2006). Because propensity to trust is a general willingness to trust others, people with different backgrounds, different cultures, and different personality types may vary in their propensity to trust levels (Mooradin et al., 2006). These individual differences may influence whether an individual trust or feels an obligation to the police. Hamm and colleagues (2017) use propensity to trust to predict trust in the original study, but ultimately find that it fails to significantly predict trust when controlling for trustworthiness. Because propensity to trust was not predictive, it was removed from the final model but this approach failed to account for mediating influence through previous stages (Hamm et al., 2017). Wee (2014) suggests that social hierarchy in Confucian values such as respect for parent’s authority leads individuals in Confucian culture to feel an obligation to obey 20 parents. As such individual differences such as social hierarchy, social harmony, and humbleness may also influence individuals’ sense of obligation to obey. Stage 4. In the fourth stage, individuals react based on their internalizations from Stage 3. Thus, if they had positive internalized feelings towards the police in Stage 3, it will likely lead to higher compliance or compliant behavior with the police. This reaction is likely to be influenced by individual differences such as propensity to trust and Confucian values. Studies have identified an association between propensity to trust and compliance (Rotter, 1971; 1980). Colquitt and colleagues (2007) found that trust was a significant mediator between propensity to trust and citizenship behavior in organizations which include compliance. Alcock and Mansell (1977) suggest that propensity to trust should lead to more cooperative behavior because individuals who have a high propensity to trust will view that others will be cooperative as well. Van Dyne and colleagues (2000) also found that propensity to trust was a significant predictor of cooperative behavior towards an organization. Confucian values may also influence cooperative and compliant behavior with the police. In Confucian societies, there is an obligation to accept a certain set of shared responsibilities, such as cooperation and compliance to social norms, as a member of the society (Phuong-Mai et al., 2005). Young and colleagues (2016) examined tax compliance and Confucian values and found that values such as maintenance of harmony, and face-saving was a strong predictor of compliance. 21 CHAPTER 4. PRESENT STUDY This dissertation examines the role of the person-level construct in predicting police legitimacy. Although there has been research conducted in South Korea regarding police legitimacy, this work has tended to focus on the relational concerns. No Asian study to date has considered propensity to trust or Confucian values in the policing context. This is important because these constructs may provide important information, especially in more ambiguous situations where there is less relational information on which to rely. The current dissertation also tests Hamm and colleagues’ (2017) Integrated Framework of Legitimacy in a novel context. As reviewed in the previous chapter, research suggests that propensity to trust and Confucian values may have important roles to play in increasing the explanatory utility of the IFL. This research therefore asks whether the person-level constructs significantly predict the constructs within the stages of the IFL, whether the person-level constructs importantly increase the variance accounted for in the IFL models, and whether there are significant indirect effects of the person-level constructs? The current dissertation, therefore, hypothesizes that propensity to trust (H1), social hierarchy (H2), social harmony (H3), face- saving (H4), and humbleness (H5) will significantly directly and indirectly predict, increase variance in constructs in all four stages of the IFL. Methods Data for the current research were collected in South Korea. South Korea is an important setting for this research for several reasons. First, prior police legitimacy research tends to ignore cultural values and only focuses on relational aspects of police-citizen encounters. For this reason, the current study specifically focuses on Confucian values. Thirdly, although recent studies regarding South Koreans’ perceptions of police legitimacy have been emerging (e.g., 22 Kim, Ra & McLean, 2019; Lee & Cho, 2019; Peak, Nalla & Lee, 2018; Seo & Lee, 2018; Woo, Maguire & Gau, 2018), more extensive investigation is needed. Therefore, there is an opportunity to apply the IFL in the South Korean context for an in-depth investigation of the perception of police legitimacy. Research Site. South Korea is a relatively homogenous society. Ninety-six percent of the population are ethnic Koreans and typically share many of the same cultural values (Kenneth, 2019; Shin & Moon, 2019). South Korea is divided into nine provinces and seven independently governed metropolises. South Korea is a highly industrialized country and 90% of the population (46 million) live in the metropolises (approximately 5 million residents live in the provinces; Robinson, 2018) which are heavily urbanized and populated. Excluding the capital, the socio- economic differences among the metropolises are relatively minor (Suh, 2019). Seoul, however, stands out from the rest of the metropolises because of its relatively higher wealth and size. Over ten million people reside in the capital and another eleven million people commute in for work (Jin & Kim, 2018). Provinces are generally less populated, but the majority are still largely industrialized. Most students come from the metropolises and less than four percent are from rural areas (Ministry of Education, 2020). The current study sampled universities from all seven metropolises (Seoul, Suwon, Daejeon University in Daejeon, Gyeonggi University in Suwon, Jeonnam University in Gwangju, Daegu University in Daegu, Ulsan University in Ulsan, and Busan University and Busan University of Foreign Studies in Busan). Universities were selected based on the first author’s professional network. Data Collection. Undergraduate students were approached by their professors during criminal justice undergraduate classes and invited to participate. Participants were informed of their rights and were assured that there would be no negative consequences for not participating 23 as survey was purely voluntary. To ensure anonymity, no identifiers were used. Participants indicated their informed consent to participate by returning the survey to the administrator. If they wish not to participate, the participants were told that they could leave the survey on their desks and not turn it in. Data collection continued until a minimum of 200 participants were recruited at each university. However, one university (Jeonnam University in Gwangju Metro) could not get 200 participants and only recruited 132. In total, data was collected from 2188 South Korean university students (see Table 1). Table 1. Description of Universities Name Metropolis Dongguk University Seoul Chungnam National University Daejeon Daejeon University Geonyang University Daegu University Daegu Ulsan Ulsan University Busan University Busan Busan University of Foreign Studies Gyeonggi University Jeonnam University Suwon Gwangju Rank #15 #26 #119 #78 #34 #35 #36 #60 #37 #38 # Students 17,986 20,914 13,908 8,256 24,540 15,834 38,492 12,352 18,659 27,054 24 Figure 2. Distribution of Provinces and Metropolises in South Korea Metropolises 1) Seoul 2) Suwon 3) Daejeon 4) Daegu 5) Gwangju 6) Ulsan 7) Busan Survey. The survey instrument was developed in English, translated to Korean, and validated by four criminal justice graduate students who are fluent in both languages. The Korean version of the survey was then back translated by a separate graduate student who had not seen the English version of the survey and evaluated by the researcher to ensure that it conveyed the same concepts. The survey elicited participants’ informed consent before presenting measures of the focal constructs. All items appeared in blocks organized by construct and started with the measures from the IFL which were presented in order by stages. Person- level constructs were measured next and the survey ended with a brief demographic questionnaire. All items were measured on one to six Likert scales without a neutral midpoint 25 unless otherwise specified (1 (Strongly Disagree), 2 (Disagree), 3 (Slightly Disagree), 4 (Slightly Agree), 5 (Agree), and 6 (Strongly Agree)). Procedural Justice (6 Items). Procedural justice was measured using six items from Hamm and colleagues (2017). The measure asked participants to consider three components of the typical behavior of law enforcement: the quality of treatment, the quality of decision-making. The questions asked participants to indicate the extent to which they agree with a series of statements regarding the police in their community. Quality of treatment items asked whether they agreed that police treat citizens with respect, police treat everyone the same, and whether the police are honest when interacting with the citizens. Quality of decision-making items asked whether they agreed that police explain their decisions to the public when asked to do so, police give citizens a chance to voice their concerns and opinions, and police behave impartially when interacting with citizens. Distributive Justice (4 Items). Distributive justice addresses the neutrality, consistency, and equality of treatment by officers. Distributive justice was measured using four items from Reisig and colleagues (2007). The questions asked whether participants agreed that police are generally consistent, they provide the same quality of services to all groups, they make citizens receive just outcomes they deserve, or if they provide better service to wealthier citizens. The original measure also included a fifth item regarding police help for minorities, but this was excluded because Korean society is a largely ethnically homogenous country. Police Effectiveness (4 Items). Police effectiveness addresses the participant’s perception of how effective the police are in handling crime in their community. Police effectiveness was measured using four items from Jonathan-Zamir and Harpaz (2014). Questions asked to what 26 extent the participants agree that police are effective in handling violent crimes, drug crimes, property crimes, and all crimes in their community. Trustworthiness and Normative Alignment (7 Items). Trustworthiness was measured using three items from Hamm and colleagues (2017). The three items measured three dimensions of trustworthiness which were ability, benevolence, and integrity. Questions asked to what extent the participants agreed that police in their community approached their job with a strong moral code, cared about the people in the community, and have the necessary skills to do their job. Additionally, normative alignment was measured using four items from Tyler and Jackson (2014). The questions asked whether the participants agreed that the participant and the police share the same sense of right and wrong, want the same things for the community, that you share similar norms and values, and that the police stand up for values that are important to the participant. Trustworthiness and normative alignment are often discussed as separate constructs but, in line with the Hamm and colleagues (2017) study, the current study conceptualizes them as one construct. Trust (3 Items). Trust measures the participant’s willingness to accept vulnerability to the police. Trust was measured using three items from Mayer and Davis (1999). Questions asked participants the extent to which they agreed that they were comfortable being vulnerable to the judgment of police, comfortable letting police handle specific situations important to them, and comfortable with the police making decisions to maintain order. Obligation to Obey (4 Items). Obligation to obey addresses an individual’s sense of obligation to obey the law. Obligation to obey the law was measured using four items from Tyler and Jackson (2014). Questions asked to what extent the participants agreed with the perception of obedience to law such as whether all laws should be strictly obeyed, if people should do what 27 the law says, whether a person who disobeys the law is a danger to the society, and if obeying the law ultimately benefits everyone in the society. Compliance (5 Items). The current study used a modified version of Sunshine and Tyler’s (2003) compliance measure. The original measure asked participants how often they followed rules about seven behaviors such as where to legally park a car, how to dispose of trash and litter, against making too much noise at night, against speeding or breaking other traffic laws, against buying possibly stolen items on the street, against taking inexpensive items from stores without paying, and against using drugs such as marijuana in public spaces. However, because of the cultural differences in the South Korean context, the questions were slightly modified. The question regarding drug use was omitted due to the relative lack of drug issues in South Korea. Not making noise at night was also replaced with a question about downloading digital contents illegally. Cooperation (4 Items). Cooperation addressed individual’s willingness to cooperate with the police. Cooperation with police was measured using four items taken from Reisig and colleagues (2007). The questions asked how likely the participant would be to call the police to report a crime, report suspicious activity near their house to the police, to report an accident that they witnessed, and provide information to help find a suspected criminal. Propensity to Trust (3 Items). The first person level construct was propensity to trust which was addresses an individual’s dispositional propensity to trust others across situations. The current study used three items from Hamm and colleagues (2013) which asked participants to report the extent to which they agreed that most people can be trusted, most people would try to be fair, and most people try to be helpful majority of the time. 28 Social Hierarchy (3 Items). Social hierarchy addresses an individual’s view of hierarchical interpersonal relationships. Social hierarchy was measured using three items from the Asian Barometer Survey (ABS, 2016). The questions asked whether participants agreed that they should always obey their parents, elders, and that students should not question their teacher’s authority. Social Harmony (3 Items). Social harmony addresses an individual’s view of maintenance of harmony and was also measured here using three items from the Asian Barometer Survey (ABS, 2016). Questions asked to what extent the participants agreed with the social harmony statement that participants should avoid open quarrels to preserve harmony; if participants would avoid conflict if they had a disagreement; and whether a person should not insist their opinion repeatedly. Face-Saving (3 Items). Face-saving addresses an individual’s view of the importance of preserving their own and their family’s face or pride. Face-saving was measured using four items from Oetzel and Ting-Toomey (2003). Questions asked if participants were concerned with shaming themselves, whether they are insecure about other’s judgment, if they were concerned with shaming their family, and if they would feel ashamed to lose their face. Humbleness (3 Items). Humbleness addressed the participant’s thoughts about not boasting and keeping one’s accomplishments in perspective. Humbleness was measured using three items from Kim and colleagues (2005). Questions asked whether participants avoided singing own praises, whether they viewed boasting is a sign of weakness and insecurity, and if they do tell their achievements if they were asked to do so. Analytic Strategy. To assess the research questions at the core of this research, the following steps are taken. First, descriptive statistics, such as frequency, range, mean, standard 29 deviation, and normality were assessed for each variable to explore the basic features of the data. Next, Confirmatory Factor Analytic (CFA) models were estimated, bivariate relations among latent constructs were tested, and reliability coefficients were computed. All models used in this study used the Maximum Likelihood (ML) estimator unless specifically stated otherwise. Global fit was assessed using Hu and Bentler’s (1999) guidelines which focus on four alternative fit indices: the Comparative Fit Index and Tucker Lewis Index (CFI, TLI; values greater than .95 are recommended), Root Mean Square Error of Approximation (RMSEA; values less than .08 are recommended), and Standardized Root Mean of the Residual (SRMR; values less than .05 are recommended). Potential changes to address poor fit were identified using modification indices which highlight item-level relations that are stronger or weaker than expected by the model. Thus, these indices can be used to identify redundant items or items that are not sufficiently related to the rest to reasonably measure the same construct, but conceptual and theoretical precision was the primary driver of all adjustments. As a result, no modifications were made if they could not be theoretically or conceptually justified. Structural equation modeling (SEM) was then used to test the hypotheses regarding significant prediction, changes in variance accounted, and indirect effects. To address the possibility that the significance of the model relations was driven by the large sample, structural models were also tested in smaller random sub-samples. An a-priori power analysis was conducted using the effect-sizes identified in the full model to determine the appropriate sample size for a smaller reanalysis (Soper, 2019; Westland, 2010). Four random sub-samples of 492, 493, 491, and 489 participants were created, and the structural model was re-estimated in each. Effects that were significant in the full data models but inconsistently 30 significant across sub-sample models were identified as suspect. All analyses were conducted in STATA version 15.1. 31 Demographics CHAPTER 5. RESULTS Data were collected from 2188 South Korean university students. 46.67% of the sample were male students (n=974) and 53.33% of the sample were female students (n=1113). The academic year of the participants were collected. Freshman (n=625) and Sophomores (n=649) were the majority of the population with 59.26% combined. Juniors (n=534) and Seniors (n=342) were 40% of the sample (M = 2.30, SD = 1.07). Age ranged from 19 to 28. Specifically, individuals between the age of 19 and 24 years old accounted for 86% of the sample, and the majority of the participants (60.51%) were between 19 and 22 years old. Furthermore, the majority of the participants (62.08%) reported that their annual household income was in the middle of 40-60% of South Korea (M = 2.87, SD = .80). Data were collected from seven metropolises in South Korea, Seoul (n=263) accounted for 12.02%, Suwon (n=227) accounted for 10.37%, Daejeon (n=642) accounted for 29.34%, Daegu (n=244) accounted for 11.15%, Gwangju (n=132) accounted for 6.03%, Ulsan (n=245) accounted for 11.20%, and Busan (n=435) accounted for 19.88% of the entire sample (M = 4.02, SD = 2.02). As provided in the data collection section, two metropolises, Busan (n=435) and Daejeon (n=642) stand out with a surplus of participants compared to other metropolises because three universities were sampled from Daejeon and two universities were sampled from Busan metro. Table 2. Demographics Academic Year Freshman Sophomore Junior Senior N 625 649 534 342 % 29.07 30.19 24.84 15.91 Min Max 1 4 Mean 2.30 SD 1.07 32 Table 2 (cont’d) Male Female Bottom 20% Bottom 40-20 % Middle 40-60% Top 20-40% Top 20% Gender Socio-Economic Status Age 19 20 21 22 23 24 25 26+ Metropolis Seoul Suwon Daejeon Daegu Gwangju Ulsan Busan 974 1113 107 423 1339 202 86 36 538 422 314 332 231 157 135 263 227 642 244 132 245 435 46.67 53.33 4.96 19.61 62.08 9.36 3.99 1.66 24.85 19.49 14.50 15.33 10.67 7.25 6.24 12.02 10.37 29.34 11.15 6.03 11.20 19.88 0 1 1 1 1 5 8 7 .53 .49 2.87 .80 4.09 1.89 4.02 2.02 Univariate Analyses Excluding the demographic variables, all items were scored on a one to six, Strongly Agree to Strongly Disagree scale. Item means were generally near the midpoint and evaluation of the standard deviations suggested that participants tended to diverge from that mean by approximately one. Normality was evaluated using Gravetter and Wallnau’s (2014) thresholds for skewness (+/-3) and kurtosis (+/-7). All study values were within these ranges, suggesting that the data are approximately normally distributed and appropriate for latent modeling (Gelman & Hill, 2007; Kline, 2016). Univariate analyses for all variables excluding the demographics variables are presented below (see Table 3. Description of Study Variables). 33 Table 3. Description of Study Variables V a l i d N Items M i s s i n g M I N M A X M e a n S t d D . S k e w n e s s K u r t o s i s PJ1: Treat with respect PJ2: Treat same PJ3: Honest PJ4: Explain PJ5: Voice PJ6: Neutral PE1: All crime PE2: Violent crime PE3: Drug crime PE4: Property crime DJ1: Consistent DJ2: Same service DJ3: Deserved outcome DJ4: Better service TWNA1: Integrity TWNA2: Benevolence TWNA3: Ability Procedural Justice 2 4 5 5 5 1 6 4.39 1.05 2186 1 6 3.98 1.29 2184 1 6 4.34 1.03 2183 1 6 4.39 1.00 2183 1 6 4.17 1.05 2183 2175 13 1 6 4.20 1.17 -0.73 0.57 -0.43 -0.43 -0.55 0.34 -0.53 0.35 -0.36 0.00 -0.56 0.05 Police Effectiveness 5 7 6 1 6 4.11 1.06 2183 1 6 4.22 1.08 2181 2182 1 6 4.13 1.13 2178 10 1 6 4.10 1.14 -0.54 0.43 -0.57 0.39 -0.62 0.33 -0.63 0.31 Distributive Justice 9 8 1 6 4.08 1.15 2179 2180 1 6 4.15 1.18 2175 13 1 6 4.08 1.15 2175 13 1 6 3.78 1.33 -0.58 0.22 -0.58 0.13 -0.51 0.11 -0.39 -0.48 Trustworthiness 2170 18 1 6 4.12 1.06 2174 14 1 6 4.34 0.99 2168 20 1 6 4.37 0.99 -0.55 0.48 -0.63 0.82 -0.71 0.93 Normative Alignment TWNA4: Right wrong TWNA5: Want same thing TWNA6: Similar values 2175 13 1 6 4.14 1.06 2177 11 1 6 4.16 1.07 2166 22 1 6 4.08 1.06 -0.53 0.34 -0.55 0.44 -0.48 0.30 Obligation to Obey Law O2OL1: Strictly obey law O2OL2: Do what the law says O2OL3: Disobey endangers society O2OL4: Obey benefit society 8 8 7 1 6 4.93 0.93 2180 1 6 4.81 0.89 2180 2181 1 6 4.78 0.97 2177 11 1 6 4.79 0.97 -0.95 1.40 -0.66 0.96 -0.58 0.17 -0.76 0.83 Trust1: Vulnerable judgment Trust2: Handle situation Trust3: Maintain order ProTrust1: People trusted ProTrust2: People fair ProTrust3: People helpful Trust 9 2179 1 6 4.14 0.89 2176 12 1 6 3.90 1.00 2177 11 1 6 4.23 0.91 -0.63 1.43 -0.45 0.48 -0.55 1.14 Propensity to Trust 2176 12 1 6 3.84 1.10 2178 10 1 6 3.95 1.06 2170 18 1 6 4.08 0.98 -0.49 0.06 -0.54 0.29 -0.57 0.79 34 Table 3. (cont’d) SocHier1: Child obeys SocHier2: Obey elders SocHier3: Obey teacher SocHier4: Vertical order SocHier5: Social standing SocHier6: Wealth status Social Hierarchy 9 2178 10 1 6 2.47 1.14 2179 1 6 2.28 1.12 2176 12 1 6 3.01 1.29 2178 10 1 6 3.97 1.04 2176 12 1 6 3.53 1.09 2171 17 1 6 4.04 1.22 -0.33 0.52 -0.26 0.62 0.12 -0.74 -0.71 0.94 -0.37 -0.02 -0.63 0.25 SocHar1: Avoid quarrel SocHar2: Avoid conflict SocHar3: No repeated opinion 2177 11 1 6 3.85 1.14 2178 10 1 6 3.45 1.17 2177 11 1 6 2.84 1.14 -0.45 0.14 -0.15 -0.33 -0.44 0.17 Social Harmony Humble1: Avoid own praises Humble2: Boasting is weakness Humble3: Only boast when asked 2177 11 1 6 3.41 1.28 2172 16 1 6 2.50 1.25 2171 17 1 6 3.58 1.23 -0.22 0.53 -0.31 -0.58 -0.47 -0.34 Humbleness Face-Saving FaceSave1: Shame myself FaceSave2: Care about others’ views FaceSave3: Protect family pride FaceSave4: Ashamed to lose face 2173 15 1 6 3.17 1.70 2175 13 1 6 3.89 1.28 2176 12 1 6 4.11 1.12 2176 12 1 6 3.84 1.22 -0.96 -0.09 -0.53 -0.18 -0.66 0.57 -0.51 0.02 Compliance with the Law Comp1: Jaywalking Comp2: Littering Comp3: Non-smoking area offense Comp4: Illegal downloading Comp5: Cheating 2175 13 1 6 3.67 1.26 2174 14 1 6 2.84 1.33 2175 13 1 6 1.95 1.33 2174 14 1 6 2.92 1.37 2175 13 1 6 1.56 1.04 -0.25 -0.66 0.16 -0.99 1.29 0.66 0.09 -1.01 2.20 4.69 Cooperation Coop1: Report crime Coop2: Report suspicious activity Coop3: Report accident Coop4: Provide information 2167 21 1 6 4.98 0.98 2168 20 1 6 4.75 1.04 2168 20 1 6 5.02 0.98 2169 19 1 6 5.01 0.99 -1.10 1.83 -0.82 0.86 -1.11 1.74 -1.03 1.33 Mean Differences Mean differences are provided in Table 4. All latent constructs examined had a minimum of one and a maximum of 6 (range: 1-6). First, the mean differences based on gender, male and female, were examined by creating item-averages for of the study constructs. Using independent means t-test, it revealed that majority of legitimacy constructs had significant differences based on gender. However, cooperation did not have any significant differences in male (M=4.93, 35 SD=.89) and female (M=4.97, SD=.87) groups; t= -1.18, p= .24. Second, socio-economic status was recoded to lower class (lower 20%; lower 40-20%; middle 40-60%) and upper class (upper 40-20% and upper 20%). Majority of the legitimacy construct did not have significant differences based on socio-economic status. However, obligation to obey and compliance did have significant differences. In particular, obligation to obey had significant differences in lower (M=4.78, SD=.77) and upper (M=4.90, SD=.86) groups; t= -2.4, p< .05; and compliance had significant differences based on lower (M=4.70, SD=.91) and upper (M=4.57, SD=1.11) groups; t= 2.19, p< .05. Lastly, mean differences were examined based on age. The younger group (19- 22) and older group (23-26) were compared and significant group differences were identified (see table 4). However, police effectiveness, trust, obligation to obey, and propensity to trust did not have significant group differences. M SD t Procedural Justice SD M Distributive Justice t SD M Police Effectiveness t 4.99*** 2.70** 4.21 4.01 4.16 4.03 1.07 .97 Trust .85 .77 4.48*** 4.29 4.03 1.04 .98 5.99*** Obligation to Obey 3.75** 4.83 4.77 .82 .74 Propensity to Trust 2.02* TWNA 4.23 4.13 4.37 4.17 Male Female Male Female Table 4. Mean Differences Gender Male Female Male Female Male Female 3.32 3.03 4.49 4.87 2.73 2.44 .97 .89 .91 .82 .97 .85 .97 .91 .99 .91 Humbleness 6.95*** Compliance Cooperation Social Hierarchy -9.40*** 4.93 4.97 .89 .87 -1.18 Social Harmony 7.13*** 3.49 3.27 .93 .84 4.04 3.89 3.58 3.68 .95 .90 3.61*** Face-saving 1.03 1.03 -2.02* 5.70*** 36 Table 4 (cont’d) M 4.11 4.05 .99 .92 SD -.02 -2.4* .86 .93 .84 2.19* -1.34 1.32 Compliance Face-saving 1.00 1.12 .77 .86 .09 .85 t t 4.26 4.23 4.18 4.14 4.15 4.14 4.78 4.90 4.09 4.09 4.94 4.93 .91 1.03 1.01 1.13 .91 1.11 .94 1.09 4.70 4.57 2.58 2.66 3.96 3.90 3.64 3.60 3.38 3.34 .88 1.03 3.18 3.09 M Humbleness Cooperation Obligation to Obey Lower Upper Lower Upper Lower Upper Lower Upper Lower Upper .93 1.01 TWNA Social Harmony Social Hierarchy Propensity to Trust 1.17 SD M Police Effectiveness .19 t 1.00 1.12 Trust .80 .90 SD M Distributive Justice .92 SD t Procedural Justice .60 SES Age Note: *p< .05, **p< .01, ***p< .001; Gray boxes represent non-significance; Range: 1-6 M Distributive Justice 19-22 23-26 19-22 23-26 19-22 23-26 19-22 23-26 19-22 23-26 t Procedural Justice M Police Effectiveness 1.59 SD Obligation to Obey -.10 Propensity to Trust .87 .99 1.05 Trust .76 .86 Social Hierarchy Social Harmony Cooperation .94 1.12 SD .91 .96 .97 .93 .87 .91 .88 .92 .91 .98 .83 .91 .78 .80 .90 .97 Humbleness 4.18 4.11 4.79 4.79 3.97 3.94 3.67 3.58 4.17 4.02 4.10 4.09 4.32 4.17 4.23 4.10 4.97 4.89 3.34 3.44 4.77 4.56 2.54 2.66 3.11 3.25 .95 .98 Face-saving 1.03 1.02 1.97* .15 2.13* 3.77*** 3.50*** -2.78** -2.72** .64 t TWNA 3.53* Compliance 5.07*** -3.38*** 1.00 1.04 Confirmatory Factor Analyses Before subjecting all of the measures to a single Confirmatory Factor Analysis, each latent construct was subjected to an individual CFA. These models aimed to assess and, if necessary, address fit at the construct level before moving to a more complicated model (see Table 5). 37 Single Factor CFAs: IFL Constructs. Procedural Justice. A CFA including only the procedural justice items fit well to the data (CFI = .96; TLI = .93; SRMR = .04; RMSEA = .14, p <.001 .00) and revealed significant loadings for all of the items above .75. Despite this good global fit, the modification indices suggested that the relations between PJ5 (“police officers give citizens a chance to voice their concerns”). This suggests that PJ5 may have been redundant with the other items and, given the central place of voice within the procedural justice literature, this makes good sense. PJ5 was, therefore, removed and a second model was estimated. The revised model also fit well to the data (CFI = .98; TLI = .97, SRMR = .02; RMSEA = .10, p < .001) and revealed factor loadings for all five items greater than .76. Reliability for the final scale was strong (Cronbach’s α = .90). Distributive Justice. The CFA for distributive justice fit well to the data (CFI = .99; TLI = .99; SRMR = .01; RMSEA = .03, p = .870) and revealed significant loadings for all of the items above .68 except DJ4 (“police provide better service to the wealthy”) which the modification indices suggested may have been less related with the other items. Specifically, DJ4 failed to load (-.23) with the rest of the items. Furthermore, because of the conceptual contradiction between DJ2 (“police provide the same service to everyone”) and DJ4 (“police provide better service to the wealthy”), the suggested removal of DJ4 make sense. Therefore, DJ4 was removed, and a second model was estimated but the deletion of DJ4 made this second model untestable (CFI =1; TLI = 1; SRMR = 0; RMSEA = 0). It, however, revealed significant loadings for all of the items above .68. Reliability for the final scale was strong (Cronbach’s α = .85). Police Effectiveness. A CFA including only the police effectiveness items fit well to the data (CFI = .97; TLI = .91; SRMR = .02; RMSEA = .02, p < .001) and revealed significant 38 loadings for all of the items above .78. Despite the good global fit, the modification indices suggested that PE1 (“police are effective in controlling all crime”) may have been redundant with the rest of the police effectiveness items. Specifically, because of the conceptual overlap of PE1 (“police are effective in controlling all crime”), PE2 (“police are effective in handling violent crime”), PE3(“police are effective in handling drug crime”), and PE4 (“police are effective in handling property crime”), removing PE1 make sense because PE1 overlaps with all three types of crime measured in PE2, PE3, and PE4. Thus, PE1 was removed, and a second model was estimated. The deletion of PE1 made this second model untestable (CFI =1; TLI = 1; SRMR = 0; RMSEA = 0). It, however, revealed significant loadings for all of the items above .85. Reliability for the final scale was strong (Cronbach’s α = .90). Trustworthiness and Normative Alignment (TWNA). Because of the conceptual similarity and consistently high correlation between the two constructs (see Hamm et al., 2017), trustworthiness and normative alignment were combined into one construct (TWNA). The modification indices suggested that TWNA2 (“police in my community care about the people in my community”) and TWNA5 (“police and I want the same things for my community”) may be less related to other TWNA items. Conceptually, TWNA2 and TWNA5 does not relate to each other. TWNA2 is regarding the benevolence of the police and TWNA5 is regarding the expectation of the normative alignment between the police and the community. Therefore, TWNA2 and TWNA5 were deleted and a second model was estimated. The second model also fit well to the data (CFI = .97; TLI = .91, SRMR = .03; RMSEA = .16, p < .001) and revealed factor loadings for all four items greater than .70. Reliability for the final scale was strong (Cronbach’s α = .85). 39 Obligation to Obey. A single factor model for obligation to obey fit moderately to the data (CFI = .96; TLI = .87; SRMR = .04; RMSEA = .18, p < .001) and revealed significant loadings for all of the items above .68. Despite the moderate fit, the modification indices suggested O2OL1 (“all laws should be strictly obeyed”) to be removed. Based on the conceptual similarity between O2OL1 and O2OL2 (“people should do what the law says”), O2OL1 was removed and a second model was estimated. The second model with O2OL2 (“people should do what the law says”), O2OL3 (“a person who disobeys the law is a danger to others in the society”), and O2OL4 (“Obeying the law ultimately benefits everyone in the society”) had improved loadings, lowest being .71. The deletion of O2OL1 made this second model untestable (CFI =1; TLI = 1; SRMR = 0; RMSEA = 0). Reliability for the final scale was acceptable (Cronbach’s α = .78). Trust. A CFA including the three trust items yielded untestable, (CFI =1.00; TLI = 1.00; SRMR = 0.00; RMSEA = 0.00) but revealed significant loadings for all of the items above .74. Reliability for the scale was strong (Cronbach’s α = .84). Compliance. A single factor model for compliance did not fit well to the data (CFI = .88; TLI = .75; SRMR = .07; RMSEA = .17, p < .001) and revealed low loadings, as low as .43, among the original five compliance items Comp1 (“how likely are you to jaywalk”), Comp2 (“how likely are you to litter”), Comp3 (“how likely are you to smoke in the non-smoking area”), Comp4 (“how likely are you to download media contents illegally”), and Comp5 (“how likely are you to cheat on exams”). The modification indices suggest that item Comp1(how likely are you to jaywalk”) to be removed. Examining the item means, while majority (69%) of the students responded that they are unlikely to jaywalk (Comp1), students responded that they are likely to litter (Comp2) (63%), smoke in a non-smoking area (Comp3) (84%), illegal download 40 (Comp4) (60%), and Cheat in school (Comp5) (92%). As such, Comp1(“how likely are you to jaywalk”) may have been less related to other items of compliance. Therefore, Comp1 was eliminated and a second model was estimated. The revised model fit well to the data (CFI = .98; TLI = .93, SRMR = .03; RMSEA = .09, p < .01) and revealed factor loadings for all four items greater than .54. Reliability for the final scale was acceptable (Cronbach’s α = .72). Cooperation. A CFA including the four cooperation items fit well to the data (CFI = .99; TLI = .98; SRMR = .02; RMSEA = .10, p < .001) and revealed significant loadings for all of the items, Coop1 (“call the police to report a crime”), Coop2 (“report a suspicious activity near your house to the police”), Coop3 (“call the police to report an accident that you witnessed”), Coop4 (“provide information to the police to help find a suspected criminal”), above .80. Reliability for the final scale was strong (Cronbach’s α = .91). Single Factor CFAs: Person-level Constructs. Propensity to Trust. A CFA including the three propensity to trust items yielded untestable, (CFI =1; TLI = 1; SRMR = 0; RMSEA = 0) but revealed significant loadings for all of the items above .76. Reliability for the scale was strong (Cronbach’s α = .87). Social Hierarchy. A CFA including only the social hierarchy items did not fit well to the data (CFI = .70; TLI = .50; SRMR = .11; RMSEA = .21, p < .001) and revealed poor loadings as low as .26. In particular, three items SocHier4 (“there is a vertical order in society”), SocHier5 (“social standings decide vertical order”), and SocHier6 (“wealth decides vertical order”), had a loading of .44, .45, and .26 respectively. Furthermore, SocHier4, SocHier5, and SocHier 6 failed to relate to rest of the social hierarchy items, SocHier1 (“even if a parent’s demands are unreasonable, children still should what their parents tell them to do”), SocHier2 (“when elder and young come into a conflict, even if the elder is in the wrong, it is better to obey the elder”), 41 and SocHier3 (“A student should not question the authority of their teacher”) which has more conceptual roots in hierarchical interpersonal relationship. Therefore, all three items, SocHier4, SocHier5, and SocHier6, were removed from the model and a second model was estimated. The new model for social hierarchy included three items and so yielded untestable, (CFI =1; TLI = 1; SRMR = 0; RMSEA = 0) but revealed factor loadings for all items greater than .54. Reliability for the final scale was moderate (Cronbach’s α = .74). Social Harmony. A CFA including the three social harmony items yielded untestable, (CFI =1; TLI = 1; SRMR = 0; RMSEA = 0) but revealed moderate loadings for majority of the items above .66 while one of the items was .46. Reliability for the scale was also moderate (Cronbach’s α = .68). Face-Saving. A CFA including the four face-saving items did not fit well to the data (CFI = .92; TLI = .75; SRMR = .05; RMSEA = .19, p < .001) and revealed loadings as low as .52. The modification indices suggested that FaceSave3 (“I am concerned with protecting the pride of my family”) was less related to the rest of the face-saving items, FaceSave1 (“I am concerned with bringing shame to myself”), FaceSave2 (“I pay a lot of attention to how others see me”), and FaceSave4 (“I feel ashamed if I lose my face”). FaceSave 1, 2 & 4 is related to saving their own reputation. Contrarily, FaceSave3 is about protecting their family’s reputation. Additionally, FaceSave3 had the lowest loading which added to the justification of its removal. The new model for social hierarchy included three items which yielded untestable, (CFI =1; TLI = 1; SRMR = 0; RMSEA = 0) but revealed factor loadings for all items greater than .59. Reliability for the final scale was moderate (Cronbach’s α = .70). Humbleness. A CFA including the three humbleness items yielded untestable, (CFI =1; TLI = 1; SRMR = 0; RMSEA = 0). The loadings for the items were, Humble1 (“I avoid singing 42 my own praises”), Humble2 (“being boastful is a sign of weakness and insecurity”), and Humble3 (“I only tell others about my achievement when I am asked to”), .63, .51, and .59 respectively. Reliability for the scale was also moderate (Cronbach’s α = .65). Multifactor CFA Having evaluated the constructs individually, I next estimated two CFA models to assess the IFL and person-level constructs separately. Integrated Framework of Legitimacy. A CFA including all eight IFL constructs fit well to the data (CFI = .96; TLI = .96; SRMR = .03; RMSEA = .04, p >.99). The CFA also revealed loadings for each latent construct (see Table 5). Most of the loadings were over .70. However, DJ3 (.68), O2OL2 (.69), Comp2 (.63), Comp3 (.67), Comp4 (.56), and Comp5 (.54) were below .70. The modification indices suggested no major issues. Correlations among the latent constructs were next assessed (see Table 6). The bivariate correlation matrix shows that the correlation among the constructs are as expected. However, compliance did not have a significant correlation with distributive justice, police effectiveness, TWNA, and trust. Items Loading .81 .84 .87 .73 .74 .82 .87 .84 .83 .87 .68 Table 5. CFA of IFL Construct Procedural Justice ( = .90) Police Effectiveness ( = .90) Distributive Justice ( = .85) PJ1: Treat with respect PJ2: Treat same PJ3: Honest PJ4: Explain PJ6: Neutral PE2: Violent crime PE3: Drug crime PE4: Property crime DJ1: Consistent DJ2: Same service DJ3: Deserved outcome 43 Table 5. (cont’d) Trustworthiness/Nor mative Alignment ( = .85) Obligation to Obey ( = .78) Trust ( = .84) TWNA1: Integrity TWNA3: Ability TWNA4: Right wrong TWNA6: Similar values O2OL2: Do what the law says O2OL3: Disobey endangers society O2OL4: Obey benefit society Trust1: Vulnerable judgment Trust2: Handle situation Trust3: Maintain order Coop1: Report crime Coop2: Report a suspicious activity Coop3: Report accident Coop4: Provide information Comp2: Littering Comp3: Non-smoking area offense Comp4: illegal download Comp5: Cheating TLI .96 All loadings were significant Cooperate with Police ( = .91) Compliance with the Law ( = .72) CFI .96 SRMR .03 .70 .76 .78 .77 69 72 71 .83 .74 .74 .81 .80 .91 .86 .63 .67 .56 .54 RMSEA .04 p >.99 Table 6. IFL Latent Construct Correlation Matrix PE TWNA TR O2O Comp Coop PJ DJ PE TWNA TR O2O Comp Coop PJ 1 .73*** .56*** .54*** .40*** .14*** .05** .17*** DJ 1 .71*** .67*** .46*** .17*** .04 1 .55*** .38*** .15*** .01 1 .42*** .16*** .03 1 .20*** .03 1 .08*** .14*** 1 .15*** 1 .18*** .17*** .17*** .16*** Note: *p< .05, **p< .01, ***p< .001; Gray colored boxes represent non-significance 44 Person-Level Constructs. A CFA for the person-level constructs (propensity to trust, social hierarchy, social harmony, face-saving, and humbleness) fit poorly to the data (CFI = .93; TLI = .91; SRMR = .05; RMSEA = .06, p < .001) and revealed significant loadings for most of the items in their latent constructs above .70 but seven items SocHier3, SocHar1, SocHar2, SocHar3, Humble1, Humble2, Humble3, and FaceSave4 were below .70 (see Table 7). The modification indices revealed no major suggestions. Lastly, I evaluated the correlations among the person-level constructs (see Table 8). The bivariate correlations were generally low, with the correlation between humbleness and face-saving being relatively high (r = .61). Humbleness and face-saving was not significantly correlated to propensity to trust. Table 7. CFA Person-Level Constructs Construct Propensity to Trust ( = .87) Social Hierarchy ( = .74) Social Harmony ( = .68) Humbleness ( = .65) Items ProTrust1: People trusted ProTrust2: People fair ProTrust3: People helpful SocHier1: Child obeys SocHier2: Obey elders SocHier3: Obey teacher SocHar1: Avoid quarrel SocHar2: Avoid conflict SocHar3: No repeated opinion Humble1: Avoid own praises Humble2: Boasting is weakness Humble3: Only boast when asked FaceSave1: Bring shame to myself FaceSave2: Care about others’ views FaceSave4: Ashamed to lose face loadings .77 .90 .82 .77 .79 .56 .67 .77 .52 .69 .56 .62 .70 .71 .58 Face Saving ( = .70) CFI .93 All loadings were significant 45 TLI .91 SRMR .05 RMSEA p < .001 .06 Table 8. Person-level Constructs Correlation Matrix ProTrst SoHier SoHarm Humble Humble ProTrst SocHier SocHarm Humble FaceSave 1 .12*** .10*** .02 -.03 1 .31*** .21*** .11*** 1 .25*** .27*** 1 .54*** 1 Note: *p< .05, **p< .01, ***p< .001; Gray colored boxes represent non-significance All Constructs. A CFA of all latent constructs in the study (the eight latent constructs from the IFL and six latent constructs from the person level constructs) fit moderately to the data (CFI = .95; TLI = .94; SRMR = .04; RMSEA = .04, p < .001) and revealed loadings for each latent constructs. All but 13 item loadings were above .70 (see Table 9). O2OL2, Comp2, Comp3, Comp4, Comp5, SocHier3, SocHar1, SocHar3, Humble1, Humble2, Humble3, FaceSave1, and FaceSave4 were below .70. Bivariate correlations were next evaluated (see Table 10). As with the smaller multiple factor CFAs, procedural justice, distributive justice, police effectiveness, and TWNA still yielded the highest correlations. The constructs of the IFL were also generally significantly related to the person-level constructs but procedural justice, distributive justice, police effectiveness, and trustworthiness and normative alignment did not significantly correlate with humbleness or face-saving. Table. 9 CFA All Latent Constructs Construct Procedural Justice ( = .90) PJ1: Treat with respect PJ2: Treat same PJ3: Honest PJ4: Explain PJ6: Neutral PE2: Violent crime PE3: Drug crime PE4: Property crime Police Effectiveness ( = .91) Items Loadings .81 .85 .86 .77 .76 .84 .87 .88 46 Table 9 (cont’d) Distributive Justice ( = .85) Trustworthiness & Normative Alignment ( = .85) Obligation to Obey ( = .82) Trust ( = .84) Cooperate with Police ( = .91) Compliance to Law ( = .72) Propensity to Trust ( = .87) Social Hierarchy ( = .74) Social Harmony ( = .68) Humbleness ( = .65) DJ1: Consistent DJ2: Same service DJ3: Deserved outcome TWNA1: Integrity TWNA3: Ability TWNA4: Right wrong TWNA6: Similar values O2OL2: Do what the law says O2OL3: Disobey endangers society O2OL4: Obey benefit society Trust1: Vulnerable judgment Trust2: Handle situation Trust3: Maintain order Coop1: Report crime Coop2: Report a suspicious activity Coop3: Report accident Coop4: Provide information Comp2: Littering Comp3: Non-smoking area offense Comp4: illegal download Comp5: Cheating ProTrust1: People trusted ProTrust2: People fair ProTrust3: People helpful SocHier1: Child obeys SocHier2: Obey elders SocHier3: Obey teacher SocHar1: Avoid quarrel SocHar2: Avoid conflict SocHar3: No repeated opinion Humble1: Avoid own praises Humble2: Boasting is weakness Humble3: Only boast when asked 47 .86 .89 .71 .77 .85 .81 .73 69 72 71 .87 .75 .79 .81 .80 .91 .86 .63 .67 .56 .54 .79 .88 .83 .77 .78 .55 .67 .77 .53 .69 .55 .61 Table 9 (cont’d) Face Saving ( = .70) CFI .95 FaceSave1: Bring shame to myself FaceSave2: Care about others’ views FaceSave4: Ashamed to lose face .68 .72 .58 TLI .94 SRMR .04 RMSEA p > .99 .04 All loadings were significant Table 10. All Latent Construct Correlation Matrix 6 1 5 2 3 4 8 9 10 11 7 1.PJ 2.DJ 3.PE 4.TWNA 5.TR 6.O2O 7.Comp 8.Coop 9.ProTrst 10.SocHi er 11.SocHa rm 12.Humb le 13.FaceS ave 1 .73* ** .56* ** .54* ** .40* ** .15* ** .06* * .17* ** .33* ** .10* ** .05* * -.02 1 .72* ** .67* ** .46* ** .17* ** .04 .19* ** .38* ** .12* ** .05* .03 -.03 -.04 1 .42* ** .16* ** .04* .17* ** .36* ** .09* ** .07* ** .03 .02 1 .55* ** .39* ** .15* ** .02 .17* ** .34* ** .11* ** .06* * .01 -.06 * 1 .20* ** .03 .16* ** .38* ** .10* ** .12* ** .08* ** .06* * 1 .08* ** .14* ** .15* ** .01 .06* ** .01 1 .14** * .04 -.16* ** -.05* * -.00 .04* -.04 1 .11** * -.13* ** -.04* * -.07* ** -.02 1 .11* ** .10* ** .01 -.02 1 .31* ** .19* ** .10* ** 1 .25* ** .26* ** Note: *p< .05, ** p< .01, *** p< .001; Gray colored boxes represent non-significance 48 12 1 3 1 1 .51* ** Structural Regressions Integrated Framework of Legitimacy. A structural equation model was next estimated to test the relationships hypothesized by the IFL in the full data. Model fit was good (CFI = .96; TLI = .96; SRMR = .04; RMSEA = .04, p > .99), all loadings were significant on their latent constructs, and the modification indices revealed little evidence of local misfit. As hypothesized and reported in Table 11, procedural justice (β = .27, p < .001), distributive justice (β = .46, p < .001), and police effectiveness (β = .22, p < .001) were all significant predictors of TWNA. TWNA was then a significant predictor of trust (β = .71, p < .001) and obligation to obey (β = .33, p < .001). Lastly, trust was a significant predictor of cooperation (β = .20, p < .001); and obligation to obey was a significant predictor of cooperation (β = .21, p < .001) and compliance (β = .15, p < .001). The variance accounted for in each criterion varied such that TWNA was largest (R2 = .79) followed by trust (R2 = .51). Obligation to obey (R2= .11), cooperation (R2 = .10), and compliance (R2 = .02) were all notably smaller. Table 11. Structural Regression IFL Direct Effects PJ TWNA TWNA DJ PE TWNA TWNA TWNA Trust Trust O2O O2O Coop Indirect Effects Trust O2O Comp Coop Comp β .27*** .46*** .22*** .71*** .33*** -.01 .20*** .15*** .21*** β Mediated by Stage 2 PJ DJ PE PJ DJ PE O2O O2O O2O Trust Trust Trust .07*** .10*** .05*** .18*** .25*** .13*** 49 SE .04 .04 .03 .02 .03 .04 .03 .03 .03 .01 .01 .01 .03 .03 .02 Table 11. (cont’d) Mediated Stage 3 Comp Coop Coop Coop Coop Comp Comp Comp .05*** .08*** .04*** .22*** .01 .01 .00 .02 .01 .01 .01 .02 PJ DJ PE TWNA PJ DJ PE TWNA R2 TWNA Trust O2O Comp Coop Model Fit CFI TLI SRMR RMSEA Note: *p< .05,**p< .01,***p< .001; Gray colored boxes represent non-significance .79 .51 .11 .02 .10 .96 .96 .04 .04 .01* .02* .01* 05* p > .99 Propensity to Trust IFL. After examining the Integrated Framework of Legitimacy, I next tested the impact of the person-level constructs on the IFL. To address the issues with the measurement of the person level constructs, models were estimated separately such that each model added one person-level construct as a predictor of each of the constructs within the IFL. The first person-level construct added was propensity to trust (see Table 12). Model fit was lower than the suggested threshold for good fit (CFI = .90; TLI = .89; SRMR = .11; RMSEA = .07, p< .001) but all loadings were significant on their latent constructs, and the modification indices revealed little evidence of local misfit. As hypothesized, propensity to trust significantly predicted Stages 1-3, procedural justice (β = .59, p < .001), distributive justice (β = .58, p < .001), police effectiveness (β = .57, p < .001), TWNA (β = .13, p < .01), trust (β = .45, p < .001), and obligation to obey (β = .18, p < .001). 50 However, propensity to trust failed to directly predict any of the constructs in Stage 4. The remaining direct effects within the IFL did not deviate much from the original model. As before, procedural justice (β = .32, p < .001), distributive justice (β = .43, p < .001), and police effectiveness (β = .25, p < .001) were all significant predictors of TWNA. TWNA was then a significant predictor of trust (β = .45, p < .001) and obligation to obey (β = .19, p < .001). Lastly, trust was a significant predictor of cooperation (β = .28, p < .001); and obligation to obey was a significant predictor of cooperation (β = .20, p < .001) and compliance (β = 15, p < .001). Compared to the original IFL model, the variance accounted for in each criterion decreased slightly but did not deviate much such that TWNA still had the biggest R2 (.74) followed by trust (R2 = .54), procedural justice (R2= .35), distributive justice (R2= .34), and police effectiveness (R2= .32). Obligation to obey (R2= .11), cooperation (R2 = .08), and compliance (R2 = .03) were all notably smaller. Regarding the indirect effects, propensity to trust had significant indirect effects on trust (β = .26, p < .001) and obligation to obey (β = .18, p < .001), mediated by Stage 2 (TWNA). Similarly, propensity to trust (β = .15, p < .001) had significant indirect effects on cooperation mediated by Stage 3(trust and obligation to obey). However, procedural justice, distributive justice, police effectiveness, and propensity to trust did not have significant indirect effects on compliance when mediated by trust and obligation to obey. In order to ensure that the significance of the effects in the model were not driven by the size of the complete sample, I next estimated the same model in four random sub-samples which revealed similar patterns. All four models revealed similar coefficients, but as would be expected, paths associated with smaller coefficients in the full sample model were non- significant in the random sub-samples models (see Table 12). Specifically, neither trust or 51 obligation to obey significantly predicted compliance in any of the sub-sample models. The sub- sample model fit statistics and the variance accounted for in each criterion were similar to the full sample model. Regarding the variance accounted in for each criterion, TWNA had the largest R2, followed by trust, and obligation to obey. The variance accounted for in cooperation, compliance, procedural justice, distributive justice, and police effectiveness were all notably smaller than in the complete sample model. IFL Sub-Sample 2 (N=493) Sub-Sample 3 (N=491) Sub-Sample 4 (N=489) .25*** .46*** .31*** .41*** .18** -.14 .29*** .10 -.18*** .60*** .58*** .56*** .09 .39*** .16* .04 .02 .21*** .14** -.07 .14** .27*** .46*** .31*** .54*** .19** -.03 .30*** .13 -.03 .52*** .53*** .52*** .14** .27*** .11 -.02 -.03 .33*** .15** .01 .17*** .35*** .48*** .28*** .39*** .14* -.07 .26*** .25*** -.11* .58*** .56*** .54*** .05 .41*** .15* -.04 .05 .24*** .10* .02 .13** Table 12. Structural Regression: Propensity to Trust Full Model (N=2,066) Sub-Sample 1 (N=492) Trust O2OP Comp Coop Comp TWNA TWNA TWNA Direct Effects (β) PJ DJ PE TWNA TWNA Trust Trust O2OP O2OP Coop ProTrst ProTrst ProTrst ProTrst ProTrst ProTrst ProTrst ProTrst Indirect Effects (β) PJ DJ PE TWNA Trust O2OP Comp Coop .32*** .43*** .25*** .45*** .19*** -.02 .28*** .15*** .20*** .59*** .58*** .57*** .13** .45*** .18*** .02 .03 Mediated by Stage 2 ProTrst ProTrst Trust O2OP .26*** .18*** Mediated by Stage3 ProTrst ProTrst Comp Coop -.02 .15*** .41*** .28*** .32*** .40*** .22** -.15 .16* .11 -.05 .56*** .57*** .59*** .13** .34*** .15* .10 .15* .25*** .16** -.06 .09 52 Table 12. (cont’d) .33 .31 .29 .74 .52 .07 .06 .09 .88 .87 .13 .08 .31 .33 .35 .73 .48 .12 .03 .07 .35 .34 .32 .74 .54 .11 .03 .08 R2 PJ DJ PE TWNA Trust O2OP Comp Coop Model Fit CFI TLI SRMR RMSEA Note: *p< .05,**p< .01,***p< .001; Gray colored boxes represent non-significance .36 .33 .32 .70 .55 .10 .02 .10 .27 .28 .27 .75 .57 .07 .02 .08 .88 .86 .12 .08 .90 .89 .11 .07 .90 .88 .13 .07 .90 .88 .12 .07 p< .001 Social Hierarchy IFL. After examining the impact of propensity to trust on the IFL, I next tested the impact of Confucian values on the IFL. The first Confucian value was social hierarchy. Model fit was lower than the suggested threshold for good fit (CFI = .87, TLI = .85, SRMR = .20, and RMSEA = .08, p< .001), but the modification indices revealed little evidence of local misfit (see Table 13). As hypothesized, social hierarchy significantly predicted all stages of the IFL in the full data, procedural justice (β = .25, p < .001), distributive justice (β = .25, p < .001), police effectiveness (β = .23, p < .001), TWNA (β = -.05, p < .01), trust (β = .12, p < .001), obligation to obey (β = .36, p < .001), compliance (β = .26, p < .001), and cooperation (β = -.25, p < .001). Relations within the IFL did not change much. Compared to the original IFL model, trust was not a significant predictor of compliance. In this model, trust was a significant predictor of cooperation (β = .31, p < .001) and compliance (β = -.13, p < .01); however, obligation to obey was not a significant predictor of cooperation. Compared to the original IFL model, the model fit and the variance accounted for in each criterion decreased. 53 Regarding indirect effects, social hierarchy had significant indirect effect on trust (β = .13, p < .001) and obligation to obey (β = .06, p < .001) when mediated by TWNA. Similarly, social hierarchy had significant indirect effects on cooperation (β = .08, p < .001) and compliance (β = -.01, p < .001) mediated by Stage 3 I next estimated the same model in four random sub-samples which revealed patterns similar to the full data. All four models revealed similar coefficients, but as would be expected, paths associated with smaller coefficients in the full sample model were non-significant in the random sub-sample models (see Table 13). Specifically, obligation to obey did not significantly predict compliance or cooperation in any of the sub-sample models. If there was a non-significant relationship in one of the sub-samples, the relationship was considered inconsistent. Trust did not consistently predict compliance and social hierarchy also did not consistently predict procedural justice, distributive justice, and police effectiveness. Also, social hierarchy was not a significant predictor of TWNA in any of the random sub-samples. The sub-sample model fit statistics and the variance accounted for in each criterion were similar to the full sample model. Table 13. Structural Regression: Social Hierarchy Sub-Sample 1 (N=492) Full Model (N=2,075) IFL Sub-Sample 2 (N=493) Sub-Sample 3 (N=491) Sub-Sample 4 (N=489) Trust O2OP Comp Coop Comp TWNA TWNA TWNA Direct Effects (β) PJ DJ PE TWNA TWNA Trust Trust O2OP O2OP Coop PJ SocHier DJ SocHier SocHier PE TWNA SocHier .36*** .53*** .39*** .59*** .18*** -.12*** .31*** .02 -.09 .21*** .20*** .21*** -.03 .36*** .56*** .42*** .64*** .19*** -.05 .27*** .06 .06 .11 .09 .07 -.02 .53*** .44*** .41*** .59*** .29*** .05 .19*** .11 .05 .12* .19*** .15** -.03 .50*** .45*** .38*** .58*** .20*** -.13*** .31*** .06 .01 .25*** .25*** .23*** -.05* .51*** .34*** .45*** .57*** .23*** -.09 .25*** -.00 .07 .23*** .21*** .15*** -.01 54 Table 13. (cont’d) Trust O2OP Comp Coop SocHier SocHier SocHier SocHier Indirect Effects (β) .12*** .36*** -.26*** -.25*** Mediated by Stage 2 SocHier SocHier Trust O2OP .13*** .06*** Mediated by Stage3 .06 .35*** -.32*** -.24*** .09*** .05** .59*** .39*** -.21*** -.21*** .09*** .05** .11* .39*** -.21** -.25*** .05 .02 .01 .01 .01 .01 .62 .43 .20 .06 .11 .87 .85 .21 .08 .12* .27*** -.21*** -.20*** .08** .05** .01 .05** .02 .04 .02 .67 .39 .19 .07 .06 .87 .85 .21 .08 Comp Coop -.02 .03 .07 -.01*** .08*** -.02*** .06** .06 .06 .05 .65 .39 .21 .08 .12 .05 .05 .02 .62 .34 .21 .010 .09 SocHier SocHier R2 PJ DJ PE TWNA Trust O2OP Comp Coop Model Fit CFI TLI SRMR RMSEA Note: *p< .05,**p< .01,***p< .001; Gray colored boxes represent non- significance .05 .04 .04 .59 .38 .22 .05 .12 .87 .85 .20 .08 .87 .85 .20 .08 .84 .82 .19 .08 p< .001 Social Harmony IFL. Model fit was lower than the suggested threshold for good fit (CFI = .86, TLI = .86, SRMR = .21, and RMSEA = .08, p< .001), but the modification indices revealed little evidence of local misfit (see Table 14). As hypothesized, social harmony significantly predicted Stages 1, 3, and 4, procedural justice (β = .17, p < .001), distributive justice (β = .15, p < .001), police effectiveness (β = .16, p < .001), trust (β = .18, p < .001), obligation to obey (β = .22, p < .001), compliance (β = .08, p < .05), and cooperation (β = -.13, p < .001). However, social harmony failed to predict TWNA throughout the full data and the sub- 55 samples. The direct effects within the IFL construct were similar to the original model. Contrarily to the original IFL Model, trust was a significant predictor of compliance (β = -.12, p < .001). The variance accounted for in each criterion decreased compared to the original IFL model. Social harmony had a significant indirect effect on trust (β = .12, p < .001) and obligation to obey (β = .06, p < .001) mediated by Stage 2 and had a significant indirect effect on cooperation (β = .08, p < .001) mediated by Stage 3. All four models revealed similar coefficients, but paths associated with smaller coefficients in the full sample model were non- significant in the random sub-sample models (see Table 14). Excluding the path between trust and cooperation, the rest of the constructs failed to have consistent predictive power. Specifically, trust and obligation to obey did not consistently predict compliance in the sub- sample models. Obligation to obey failed to predict cooperation in all four sub-samples. Social harmony failed to consistently predict all constructs except obligation to obey in the sub- samples. Social harmony failed to predict TWNA in both the full sample and the sub-samples. The variance accounted for in each criterion decreased compared to the original IFL model. Table 14. Structural Regression: Social Harmony Sub-Sample 1 (N=493) Full Model (N=2,077) IFL Sub-Sample 2 (N=497) Sub-Sample 3 (N=492) Sub-Sample 4 (N=486) TWNA TWNA TWNA Direct Effects (β) PJ DJ PE TWNA TWNA Trust Trust O2OP O2OP Coop SocHarm PJ DJ SocHarm SocHarm PE Trust O2OP Comp Coop Comp .50*** .46*** .38*** .57*** .23*** -.12*** .30*** .14*** -.06* .17*** .15*** .16*** .51*** .34*** .45*** .53*** .24*** -.12* .28*** .09 -.01 .16* .15* .16* 56 .36*** .56*** .41*** .64*** .19*** -.04 .27*** .13* -.01 .07 .06 .08 .45*** .54*** .35*** .54*** .15*** -.08 .29*** .24*** -.08 .25** .19* .18* .54*** .45*** .42*** .60*** .31*** -.09 .18*** .15** .00 -.01 .01 .03 Table 14. (cont’d) TWNA Trust O2OP Comp Coop SocHarm SocHarm SocHarm SocHarm SocHarm Indirect Effects (β) .02 .18*** .22*** -.08* -.13*** Mediated by Stage 2 SocHarm SocHarm Trust O2OP .12*** .06*** Mediated by Stage3 .04 .22*** .19** -.15* -.14* .11** .07** .07 .08 .17** -.02 -.16** .08* .04* -.02 .17** .28*** -.01 -.10 .14** .05* -.02 .10** .02 .04* -.00 .08*** Comp Coop .05 .06* .03 .02 .03 .64 .41 .12 .03 .09 SocHarm SocHarm R2 PJ DJ PE TWNA Trust O2OP Comp Coop Model Fit CFI TLI SRMR RMSEA Note: *p< .05,**p< .01,***p< .001; Gray colored boxes represent non-significance .03 .02 .02 .62 .12 .39 .04 .07 .01 .00 .01 .63 .08 .43 .02 .08 .07 .04 .03 .66 .12 .37 .06 .08 .87 .85 .21 .08 .86 .86 .21 .08 .84 .82 .19 .09 .86 .84 .20 .08 p< .001 .08 .12* .15** -.12* -.09 .05 .03 .01 .04* .00 .00 .00 .68 .13 .38 .04 .04 .86 .85 .21 .08 Face-saving IFL. Model fit was lower than the suggested threshold for good fit (CFI = .86, TLI = .85, SRMR = .21, and RMSEA = .08, p< .001), but the modification indices revealed little evidence of local misfit (see Table 15). Face-saving significantly predicted only TWNA (β = .11, p < .001) and trust (β = .07, p < .01). The direct effects within the IFL construct did not deviate much from the original model. Compared to the original model of IFL, trust was a significant predictor of compliance (β = -.10, p < .001). The variance accounted for in each criterion changed compared to the original model of the IFL. TWNA (R2 = .64) still had the 57 biggest variance accounted for but obligation to obey (R2= .37) became the second biggest variance accounted for instead of trust (R2 = .07). Regarding indirect effects (see Table 15), face- saving failed to have a significant indirect effect on trust and obligation to obey mediated by Stage 2 (TWNA). Also, face-saving failed to have a significant indirect effect on compliance and cooperation via Stage 3 (trust & O2O). All four random sub-samples revealed similar coefficients, but paths with smaller coefficients in the full sample model were not significant in the sub-samples. Compared to the original IFL model, obligation to obey failed to consistently predict compliance and cooperation in the sub-samples. Trust failed to consistently predict compliance in the sub-samples. Excluding the relationship between face-saving and TWNA, face-saving failed to predict any construct consistently in the sub-samples Table 15. Structural Regression: Face-saving IFL Sub-Sample 2 (N=496) Sub-Sample 3 (N=492) Sub-Sample 4 (N=494) .36*** .57*** .43*** .66*** .22*** -.04 .26*** .14* -.04 -.15* -.17* -.19** .10* .05 .06 -.08 .06 .47*** .55*** .37*** .57*** .19*** -.08 .26*** .23*** -.08 -.08 -.08 -.12 .11* .07 .11* .04 -.17** .53*** .44*** .43*** .60*** .32*** -.08 .18*** .16** -.00 .02 .03 -.02 .12** -.01 -.02 .06 -.05 Full Model (N=2,074) Sub-Sample 1 (N=494) Trust O2OP Comp Coop Comp TWNA TWNA TWNA Direct Effects (β) PJ DJ PE TWNA TWNA Trust Trust O2OP O2OP Coop PJ FaceSave DJ FaceSave FaceSave PE FaceSave TWNA Trust FaceSave O2OP FaceSave FaceSave Comp Coop FaceSave .52*** .46*** .40*** .60*** .26*** -.10*** .28*** .15*** -.08** -.04 -.04 -.06 .11*** .07** .04 .02 -.02 .37*** .54*** .40*** .60*** .24*** -.11* .29*** .10 -.18** -.06 -.06 -.09 .11*** .09 .10 .05 .03 58 Table 15. (cont’d) Indirect Effects (β) Mediated by Stage 2 FaceSave FaceSave Trust O2OP .03 .02 Mediated by Stage3 .01 .01 -.07 -.03 Comp Coop -.00 .01 .00 -.01 -.00 .02** .00 .00 .01 .59 .07 .37 .02 .10 .00 .00 .00 .64 .07 .37 .03 .08 FaceSave FaceSave R2 PJ DJ PE TWNA Trust O2OP Comp Coop Model Fit CFI TLI SRMR RMSEA Note: *p< .05,**p< .01,***p< .001; Gray colored boxes represent non- significance .02 .03 .04 .63 .05 .43 .03 .07 .87 .85 .21 .08 .86 .85 .21 .08 .86 .84 .21 .08 p< .001 -.01 -.00 .02 .01 .01 .01 .01 .65 .05 .33 .06 .10 .86 .84 .21 .08 .07 .05 -.00 .01 .00 .00 .00 .68 .10 .36 .03 .03 .86 .85 .21 .08 Humbleness IFL. Model fit was lower than the suggested threshold for good fit (CFI = .86, TLI = .85, SRMR = .20, and RMSEA = .07, p< .001), but the modification indices revealed little evidence of local misfit (see Table 16). Humbleness significantly predicted only trust (β = .13, p < .001) obligation to obey (β = .15, p < .001), and cooperation (β = -.13, p < .001). The direct effects within the IFL were largely similar to the full data model. Compared to the original IFL model, the strength of procedural justice significantly predicting TWNA nearly doubled. Also, while trust fails to predict compliance in the IFL model, here, trust is a significant predictor of compliance (β = .59, p < .001). In the IFL model, the variances accounted for provided that TWNA and trust were the two biggest variances accounted for. However, in 59 this model, TWNA and obligation to obey were to two largest variances accounted while the variance of trust significantly dropped. As for the indirect effects, humbleness failed to have significant indirect effects on obligation to obey and trust when mediated by Stage 2 (TWNA). Also, humbleness failed to have significant indirect effects on cooperation and compliance when mediated by Stage 3 (trust & O2O). The four random sub-samples revealed patterns very similar to the full data. Humbleness failed to predict any constructs consistently across all sub-samples. The sub-sample model fit statistics and the variance accounted for in each criterion were similar to the full sample model. Table 16. Structural Regression: Humbleness IFL Full Model (N=2,070) Sub-Sample 1 (N=492) Trust O2OP Comp Coop Comp TWNA TWNA TWNA Direct Effects (β) PJ DJ PE TWNA TWNA Trust Trust O2OP O2OP Coop PJ Humble DJ Humble PE Humble Humble TWNA Trust Humble O2OP Humble Comp Humble Humble Coop Indirect Effects (β) .51*** .47*** .39*** .59*** .26*** .59*** .28*** .15*** -.06* -.01 .03 .00 .03 .13*** .15*** -.00 -.13*** Mediated by Stage 2 Humble Humble Trust O2OP .02*** .01*** Mediated by Stage3 Humble Humble Comp Coop .01*** .03** .53*** .36*** .45*** .58*** .29*** -.09 .24*** .13* -.01 -.10 -.09 -.08 .02 .19*** .26*** -.08 -.11 -.04 -.02 .02 .02 60 Sub-Sample 2 (N=493) Sub-Sample 3 (N=495) Sub-Sample 4 (N=494) .36*** .54*** .39*** .61*** .24*** -.13* .29*** .09 -.16** -.08 -.03 -.06 -.01 .15** .20** .01 -.05 -.03 -.02 .00 -.00 .50*** .49*** .44*** .60*** .31*** -.08 .18*** .17** -.01 .02 .08 .03 .06 -.02 .10 -.07 -.02 .04 .03 .01 .01 .56*** .54*** .32*** .64*** .26*** -.14** .25*** .22** .02 -.11 -.09 -.06 -.02 .20*** .26*** .06 -.22*** -.08 -.04 .03 .03 Table 16. (cont’d) .01 .01 .01 .61 .14 .36 .02 .06 .00 .00 .00 .64 .09 .38 .03 .09 R2 PJ DJ PE TWNA Trust O2OP Comp Coop Model Fit CFI TLI SRMR RMSEA Note: *p< .05,**p< .01,***p< .001; Gray colored boxes represent non-significance .01 .00 .00 .58 .09 .38 .02 .10 .00 .01 .00 .69 .11 .36 .03 .03 .84 .82 .20 .08 .86 .85 .20 .07 .86 .84 .22 .08 .85 .84 .21 .08 .01 .01 .00 .73 .12 .41 .06 .10 .85 .83 .21 .08 p< .001 61 CHAPTER 6. DISCUSSION Since Weber’s (1968) work on legitimacy, scholars have continued to debate the sources of legitimacy and the organization of legitimacy. Extensive empirical research on the Process- Based Theory of Police Legitimacy attests to the importance of police officer behavior and its impact on the perception of police legitimacy. However, prior research typically failed to examine the role of person-level constructs. Rather, the literature tends to focus on the relational aspects of police-citizen encounter as the primary source of voluntary compliance. The current dissertation sought to understand the impact of person-level constructs on police legitimacy, focusing specifically on propensity to trust and Confucian cultural values. To address this, the current dissertation used Hamm and colleague’s (2017) Integrated Framework of Legitimacy as a theoretical framework to organize the constructs related to police legitimacy. This exploratory research hypothesized that that propensity to trust (H1), social hierarchy (H2), social harmony (H3), face-saving (H4), and humbleness (H5) will significantly directly and indirectly predict, and increase variance accounted in constructs in all four stages of the IFL. Key Findings Propensity to Trust. Propensity to trust directly and significantly predicted the constructs of Stage 1 (procedural justice, distributive justice, & police effectiveness), Stage 2 (TWNA), and Stage 3 (trust & obligation to obey). However, it failed to directly predict Stage 4 (compliance & cooperation). Propensity to trust is a predisposed tendency that makes individuals more willing to trust others, regardless of the specific characteristics or behavior of the individual being trusted (Mayer et al., 1995; Frazier et al., 2013). As hypothesized, propensity to trust significantly impacted how people internalized their feelings towards the police in Stage 3. Specifically, propesnsity to trust was a significant predictor of both trust and obligation to obey 62 but the relationship between trust and propensity to trust was much stronger. Individuals with higher propensity to trust are more likely to report trust in all targets and therefore should feel more trust toward the police. This explains why propensity to trust predicting trust was one of the stronger relationships in the model. Regarding the indirect effect of propensity to trust, when mediated by stage 3 (trust & obligation) indirectly significantly predicted cooperation but had no indirect effect on compliance. As hypothesized, individual differences had a significant impact on how individuals evaluated information regarding their encounters with the police (Stage 1). One explanation for this could be that individuals with higher propensity to trust levels are more trusting and optimistic in viewing the police which may result in more positive interactions. Different people may evaluate interactions differently even when they experienced similar events. For example, research on the interpretation of ambiguous information has shown that people interpret the same information differently (Mogg, Bradbury & Bradley, 2006). Negative and positive interpretations of information were assessed and found that dysphoric (Lawson & MacLeod, 1999) and depressed (Butler & Mathews, 1983) individuals were more likely to interpret ambiguous information as negative. As such, individuals who experienced (direct or vicarious) police officer may have a different interpretation of the interaction. Individuals with higher propensity to trust may have more optimistic view in general because of their general tendency to trust others and interpret the police interaction as a positive interaction. On the other hand, individuals with low propensity to trust may negatively interpret the interaction with the police regardless of the police treatment. Similar to Alarcon and colleagues (2016) who found that propensity to trust was a significant predictor of trustworthiness, this study also found that individual differences in 63 propensity to trust significantly impacted how people evaluate the police in Stage 2. One explanation is, because propensity to trust is a general tendency to trust, it may have positively impacted the evaluations regarding the officer’s integrity (TWNA1), benevolence (TWNA2), and sense of morality (TWNA3). In organizational context, individuals who are more likely to trust in general were more likely to view their supervisor as trustworthy (Nambudiri, 2012). As hypothesized, propensity to trust had a significant indirect effect. When mediated by stage 2 (TWNA), propensity to trust indirectly significantly predicted trust and obligation to obey Contrary to what was hypothesized, propensity to trust did not have a direct significant impact on Stage 4. One possible explanation could lie in the fact that Stages 1,2 and 3 are about the individual’s perception of the police while Stage 4, the reaction stage, involves individuals acting in certain ways such as cooperating or complying with the police. Individual’s general tendency to trust others may have influenced how they perceive and evaluate the police, but this general trust tendency may have not been enough to alter their reactive behavior towards the police. Propensity to trust, however, did have a significant indirect effect on cooperation. Although propensity to trust fails to directly predict compliance and cooperation, it did influence them via Stage 3 (trust & obligation to obey) constructs. Lastly, and contrary to Hypothesis 1, the variance accounted for in the IFL constructs did not increase as compared to the original IFL model. This suggests that propensity to trust does not add any new variance to the model. Thus, this study does not refute the Process-Based Theory of Police Legitmacy’s primary focus on relational concerns (Tyler, 1990). As the analyses in this study demonstrated, the constructs within the IFL had the most significant impacts and, even though the person level constructs were predictive, they did not account for a new variance. Thus, although propensity to trust is an important factor predicting the legitimacy 64 constructs (procedural justice, distributive justice, police effectiveness, TWNA, trust & obligation to obey), the relational aspect of police legitimacy appears to dominate the models. Figure 3. Structural Regression Paths: Propensity to Trust IFL Social Hierarchy. Social hierarchy directly and significantly predicted the constructs of Stage 1 (procedural justice, distributive justice, & police effectiveness), Stage 2 (TWNA), Stage 3 (trust & obligation to obey), and Stage 4 (compliance & cooperation). As hypothesized, individual differences had a significant impact on how individuals evaluated information regarding their encounters with the police (Stage 1). One explanation for this could be that individuals with higher social hierarchy levels are more likely to respect authority. Koreans who value social hierarchy have pre-existing respect for authority which may translate to respect for the police. This may explain why social hierarchy significantly and directly impacts Stage 1. It 65 may be that citizens’ preset respect towards authority makes them interact with the police in a more positive manner. However, we cannot be sure if respect causes a positive interaction with the police. Nevertheless, it is clear that individuals with higher levels of social hierarchy are more likely to evaluate information about the police positively in Stage 1. Although Stage 2 is also significantly predicted by social hierarchy, the magnitude of the effect is much smaller. Therefore, this significance may have been due to the large sample size. Indeed, when examining the sub-samples, social hierarchy was never a significant predictor of TWNA. As hypothesized, individual differences in social hierarchy had a significant impact on how individuals internalized the evaluated information regarding their encounters with the police (Stage 3). The relationship between social hierarchy and obligation to obey makes sense. An individual who is more likely to believe in interpersonal hierarchical relations is going to feel more sense of obligation to obey the police because the police are an authority and so should be respected and obeyed. In Confucianism, individuals view their parents or teachers to be hierarchically above them. This means that individuals obey them and respect them not because they are kind, but because they are to be respected based on the social hierarchy. Police are also authorities, thus, individuals who value social hierarchy respect and feel a sense of obligation to obey them because they are hierarchically above an everyday citizen. Regarding the indirect effects of social hierarchy, social hiearchy had a significant indirect effect on trust and obligation to obey when mediated by Stage 2 (TWNA). Although the strength of the indirect effect is not large, three out of the four sub-samples supported this effect. Interestingly, and contrary to the hypothesis, individuals with higher social hierarchy values were less likely to comply and cooperate with the police (Stage 4). One explanation of 66 this may be because of rapid westernization and the socio-cultural change in South Korea. In the last three decades, South Korea has moved from a traditional agrarian society to a predominantly young, urban, educated industrialized society (Hyun, 2001). Also, because of this rapid industrialization, it has fostered new ideologies like Western liberalism where individualism is emphasized (Hyun, 2001). There is no evidence to suggest that South Korea is not a Confucian society anymore because of westernization. However, it is possible that the younger generations of South Koreans no longer support the specific Confucian ideologies such as social hierarchy that conflicts with their new Korean culture. The younger generations may no longer support the idea of blindly complying and cooperating with the authorities. Looking at the frequencies of social hierarchy, 90% of the Korean students reported that they are unlikely to agree with social hieararchy beliefs and less than 10% of the students somewhat agreed to social hiearchy beliefs. The indirect effects of social hierarchy, when mediated by Stage 3 (trust & obligation to obey), had a significant indirect effect on compliance and cooperation. With the addition of social hierarchy to the IFL model, the variance accounted for decreased. Therefore, it appears that, contrary to the hypotheses, social hierarchy does not increase the explanatory power of the IFL. 67 Figure 4. Structural Regression Paths: Social Hierarchy IFL Social Harmony. Social harmony directly and significantly predicted the constructs of Stage 1 (procedural justice, distributive justice, & police effectiveness), Stage 3 (trust & obligation to obey), and Stage 4 (compliance & cooperation). However, failed to predict Stage 2 (TWNA) across all data set including the sub-samples. Although social harmony predicted Stages 1,3, & 4 directly and significantly in the full model, the relationships were not strong. Furthermore, excluding obligation to obey, none of the relationships were consistently significant across the sub-samples. Regarding obligation to obey, however, individuals who believed in the importance of maintenance of harmony in relationships and the society were more likely to feel a sense of obligation to obey. This makes sense because the focus of social harmony is to follow social norms and maintaining harmony. Obeying parents, teachers, and authorities is a social norm of Confucianism. Thus, when an individual disobeys the police, they are disrupting social harmony. Social harmony did have indirect significance to trust and obligation to obey when mediated by Stage 2 (TWNA) and cooperation when mediated by Stage 68 3 (trust & obligation to obey). The variance accounted for did not increase when social harmony was introduced to the IFL model. Thus, it seems that, contrary to the hypotheses, accounting for social harmony generally does not increase the explanatory power of the IFL. Figure 5. Structural Regression Paths: Social Harmony IFL Face-saving. Face-saving failed to directly and significantly predict the constructs of Stage 1 (procedural justice, distributive justice, & police effectiveness), Stage 3 (trust & obligation to obey), and Stage 4 (compliance & cooperation). However, it did predict Stage 2 (TWNA). One possible explanation for failing to predict Stage 1of the IFL could be because face-saving is about maintaining own face in the view of others while constructs within Stage 1 (procedural justice, distributive justice, & police effectiveness) are about evaluations of interactions with the police. Therefore, there may have been no conceptual link between face- saving and constructs within Stage 1. As evident in the latent construct correlation matrix (see Table 10), the correlations between face-saving, procedural justice and distributive justice, are 69 insignificant while the strength of the correlation between face-saving and police effectiveness is very small and negative. Regarding Stages 3 and 4, the current study expected a direct and significant relationship between face-saving, obligation, trust, compliance, and cooperation. If individuals are worried about their reputation, they should feel a sense of obligation to obey, trust, comply and cooperate with the police to save face. To disrupt the social harmony by disobeying and not complying with the police is shameful. However, as the results show, obligation to obey, trust, compliance, and cooperation was not significantly predicted by face-saving. Additionally, there were no significant indirect effects of face-saving. Also, the variance accounted for did decrease compare to the original IFL model, meaning there was no new variance to explain the model when face- saving was introduced. Thus, contrary to the hypotheses, face-saving does not increase the explanatory power of the IFL. Figure 6. Structural Regression Paths: Face-Saving IFL 70 Humbleness. Humbleness failed to directly and significantly predict the constructs of Stage 1 (procedural justice, distributive justice, & police effectiveness), Stage 2 (TWNA), and Stage 4 (compliance & cooperation). However, did predict Stage 3 (trust & obligation to obey) but the significance was not supported across the sub-samples. The current study expected a meaningful relationship with all four stages of the IFL. Humbleness emphasizes conforming to the social norms, and not disrupting the social norm. Therefore, when an individual is humble, one should act according to the social norms and try to preserve harmony. Contrary to the hypothesis, humbleness failed to predict Stage 1 (procedural justice, distributive justice, & police effectiveness) of the IFL. One possible explanation could be because of the modernization or Westernization of South Korea for the past three decades, younger generations have fostered a new Korean culture that puts more emphasis on liberalism and individualistic culture (Hyun, 2001). As such, the younger South Koreans may no longer identify with humbleness which may have caused humbleness to be a non-significant predictor of legitimacy constructs. As evident in the latent construct correlation matrix (see Table 10), the correlations between humbleness and procedural justice, distributive justice, and police effectiveness are insignificant. When humbleness is introduced into the IFL model, the variance accounted for decreases which may indicate that it has no significant impact in the IFL. 71 Figure 7. Structural Regression Paths: Humbleness IFL Limitations The current study sought to test the role of person-level constructs in public perceptions of police legitimacy. Although it lends important insight into this question, the study is not without limitations. First, this study is limited by its use of a convenience sample of South Korean university students. The convenience sample recruited and analyzed does not represent the entire South Korean university student populations. However, because nearly 70% of South Koreans receiving tertiary education (highest percentage worldwide) after high school (Mani, 2018), assessing university students may be a reasonable representation of the younger generation of South Koreans. Additionally, the current dissertation’s sample is very similar to the average South Korean university student population. According to Korean Higher Education Statistics Survey, the total number of students enrolled in universities were 2,001,643 in 2019, the gender ratio of male to female students were 56:44, and distribution of academic class was 30: 28:25:17 (Korean Higher Education Statistics Survey, 2019). 72 Albeit convenience sampling was used, the current study’s data does not differ drastically with the average South Korean university student populations in terms of the distribution and diversity. The gender ratio was near 50:50 in this sample and the academic year distribution was near the norm with a similar number of students in each class. Socio-economic status was very similar to the average South Korean university student populations where the majority of the students reported they belonged in the middle class and very few reporting that they belong in the top 20% (4%) and bottom 20% (5%). In terms of regional diversity, the current sample does not have a region identifier to assess where the students are originally from, such as rural or metropolis. However, the majority (90%) of the South Korean population resides in the metropolis and only 10% of the population resides in the rural areas (Suh, 2019). Furthermore, the ministry of education reports that less than 4% of university students come from rural areas (Won, 2019). Therefore, it seems safe to assume that albeit covinience sample was used, the current sample does not deviate too much from the average South Korean university student population. However, the age distribution of the current sample may not be similar to the average South Korean university student populations. Out of the total, only 1.66 percent (36 students) were 19 years of age. One thing to note here is that there was 625 Freshman but only 36 students who are 19 years old. One possible explanation is that in South Korea, students can wait one year and retake the college entrance exam to better their chance of getting a higher score next year. Therefore, the majority of the Freshman students may have retaken their entrance exam which may explain 25% of the students are 20 years old. Yet, only 1.66 percent of students being 19 years old still does seem different compared to the average South Korean university student population. Another possible generalizability issue presented in this study relates to the major of 73 the students. The participants were criminal justice major students which may have some influence on how they view the police. Graduates with a criminal justice degree aim to become a police officer, corrections officer, or civilian police employees. Thus, criminal justice students may be more pro-police and are less likely to espouse especially negative views of the police. Therefore, the construct such as procedural justice, distributive justice, police effectiveness, TWNA, obligation to obey, compliance, and cooperation may have been influenced by pro- police attitudes. Second, the current study had noteworthy measurement issues. Similar to Monkhouse and colleagues’ (2013) measures of face-saving and humility, the current study also had Confucian values such as humbleness, face-saving, and compliance that suffered low to mediocre reliability and low loadings. This may mean that humbleness, face-saving, and compliance were not unidimensional. This would mean that items that were intended to measure a single latent construct addressed different subconstructs. Previous research on Confucian values has tended to struggle with identifying and agreeing with the dimensions of Confucianism and creating a reliable measure of the cultural values which caused confusion in how to measure Confucianism (Monkhouse et al., 2013). In the current research, the measures of social hierarchy and social harmony were generally strong, but the remaining constructs were significantly more challenged. This matters for the current research because, as Lebel and Paunonen (2011) suggest, “lower levels of reliability are associated with decreasing probabilities of detecting a statistically significant effect” (p.573). The current dissertation borrowed existing measures of Confucianism. However, as the CFA on person-level constructs show, certain Confucian values measures such as face-saving and humbleness lacked unidimensionality and reliability. Items of face-saving and humbleness loaded onto other constructs. Thus, although there is a good reason 74 to expect that the face-saving and humbleness would not conceptually relate to legitimacy, it is also possible that the lack of association here was the result of relatively high levels of statistical noise. Additionally, the literature on Confucian values may be outdated. Rapid modernization or Westernization has encouraged the younger generation of South Koreans to foster new social ideologies similar to Western liberalism (Hyun, 2011). Furthermore, some scholars argue that because of the rapid modernization, certain Confucian values are eroding quickly which makes it unfeasible to examine (Keller & Kronstedt, 2005; Rowley & Warner, 2010). Thus, in the future, researchers should identify whether specific Confucian values such as humbleness and face- saving is worth measuring in Confucian societies Third, the current study collected data cross-sectionally. Thus, it is not possible to determine causality because of the lack of temporal order between constructs. However, despite these limitations, the current study contributed to the police legitimacy literature because it tested potential new predictors of police legitimacy such as propensity to trust and person-level constructs. Conclusion The current dissertation examined five person-level constructs--propensity to trust, social hierarchy, social harmony, face-saving, and humbleness--to see if individual differences had any direct and indirect impact on the stages of the IFL. However, as discussed above, only two person-level constructs had a meaningful impact on police legitimacy. As the results of this study provided, although some person-level constructs had a significant impact on the IFL, the original IFL constructs were consistently more important in the full model and the four random submodels. This was evident by assessing the strength of the structural regression coefficients and the variances accounted for in the model. Even with the introduction of person-level 75 constructs in the model, the original IFL’s constructs were consistently stronger predictors of legitimacy. Although the coefficients of the original IFL do decrease slightly with the introduction of person-level constructs in the model, it does not drastically fluctuate. When assessing the variances accounted for, none of the person-level constructs increased the variances accounted for compared to the original IFL. This suggests that the introduction of person-level constructs in the model did do not explain more than the original IFL. This also suggests that, as in western cultures, South Koreans seem to link police legitimacy constructs in ways similar to Western participants. Thus it seems that the effects of procedural justice, distributive justice, police effectiveness and trustworthiness/normative alignment are not displaced by person-level constructs such as Confucian values. Although some of these person-level constructs matters for police legitimacy, the relational aspect still seems to be more important. 76 APPENDICES 77 APPENDIX A. Survey Instrument English Research Participant Information and Consent Form Thank you for participating in this survey. The aim of this research is to add to the knowledge of police legitimacy literature. Specifically, how Koreans view police authority and how cultural perspectives may explain why people obey the police. Please fill the questionnaire by choosing one option for each questions provided below. The survey administrator does not know the names of the respondents that choose to participate in the study nor does this survey have identification marks. All responses are completely anonymous and will not be used in any ways that may identify the respondent. Your privacy will be protected to the maximum extent allowable by law. Please return the questionnaire to the administrator once completed. Confidential agreement Participation in this survey is voluntary and it will take about 20 minutes. Even after you agree to respond to the survey, you may refuse to participate in certain procedures, answer certain questions, or discontinue your participation at any time without penalty or loss of benefits. To keep this survey anonymous, please do not provide any identifiable information on the form. I do not foresee any risks for responding to this survey. The submitted questionnaires will be coded into data files and both the questionnaires and the coded data will be kept with the investigator at Michigan State University (MSU) (Baker Hall, 655 Auditorium Road, Room 134, East Lansing, MI 48824) for a period of 3 years after the project closes. The data files will only be shared with the Institutional Review Board (IRB) at MSU. Coded data will be password-protected and the computers storing data will be located in the investigator’s office that is locked when unoccupied. Contact Information for Questions and Concerns If you have questions or concerns about the research, such as scientific issues, how to do any part of it, or to report an injury (i.e. physical, psychological, social, financial, or otherwise), please contact Sung Lee (Baker Hall, 655 Auditorium Road, Room 139, East Lansing, MI 48824, leesungu@msu.edu, 517-353-5150). If you have questions or concerns about your role and rights as a research participant, would like to obtain information or offer input, or would like to register a complaint about this study, you may contact, anonymously if you wish, Michigan State University’s Human Research Protection 131 Program (Phone: 517-355-2180; Fax: 517-432-4503; e-mail: irb@msu.edu; mail: Olds Hall, 408 W. Circle Drive, Room 207, East Lansing, MI, 48824). You indicate your voluntary agreement to participate by completing and returning this questionnaire. 78 This Section is regarding your perception of procedural justice of the police. Please select one between the range of 1 ~ 7. 1= Strongly Disagree 2= Disagree 3= Slightly Disagree 4= Neutral 5= Slightly Agree 6= Agree 7= Strongly Agree No. Procedural Justice 1.1 Police treat citizens with respect Content 1.2 Police treat everyone the same 1.3 Police are honest when interacting with citizens 1.4 Police explain their decisions to the public when asked to do so 1.5 Police give citizens a chance to voice their concerns and opinions 1.6 Police behave neutrally when interacting with citizens 1.7 Police overstep boundaries of their authority 1.8 Police act as if they are above the law 1.9 Police bother citizens for no good reason 1 2 3 4 5 6 7 This section is regarding your perception of the distributive justice of the police. Please select one between 1 ~ 7. No. Distributive Justice 2.1 Police enforce the law consistently when Content dealing with people 2.2 Police provide the same quality of services to all citizens 2.3 Police make sure citizens receive the outcome they deserve under the law 2.4 Police provide better service to politicians and people in power 2.5 Police provide better service to wealthier citizens 1 2 3 4 5 6 7 79 This section regards your perception of police efficiency. Please select one between 1 ~ 7. No. Police Performance 3.1 Police are efficient in handling crime in Content my community 3.2 Police are efficient in handling violent crimes in my community 3.3 Police are efficient in handling drug crimes in my community 3.4 Police are efficient in handling property crime in my community 1 2 3 4 5 6 7 This section regards your perception of police trustworthiness and normative alignment. Please select one between 1 ~ 7. No. Trustworthiness 4.1 Police in my community approach their Content job with a strong moral code 4.2 Police in my community care about the people in my community 4.3 Police in my community have the skills necessary to do their job Normative Alignment 4.4 Police generally have the same sense of right and wrong that you do 4.5 You and the police want same things for your community 4.6 The values of most officers who work in your community are similar to your own 1 2 3 4 5 6 7 80 This section regards your perception of obligation to obey the police and the law. Please select one between 1 ~ 7. No. Obligation to Obey the Law 5.1 All laws should be strictly obeyed Content 5.2 People should do what the law says 5.3 A person who disobeys the law is a danger to others in the society 5.4 Obeying the law ultimately benefits everyone in the society Obligation to Obey the Police 5.5 You should support the decisions of the police even when you disagree with them 5.6 You should do what the police tell you to do even if you do not understand or agree with the reasons 5.7 You should do what the police tell you to do even if you do not like how they treat you 5.8 Police are legitimate authorities; therefore, you should do what they tell you to do 1 2 3 4 5 6 7 81 This section regards your opinion of trust related topic. Please select one between 1 ~ 7. Content 1 2 3 4 5 6 7 No. Trust 6.1 6.2 6.3 I am generally comfortable being vulnerable to the judgment of the police I would be comfortable letting police handle a specific situation that is important to me I am generally comfortable with police making decisions to maintain order Propensity to Trust 6.4 Generally speaking, I would say that most 6.5 6.6 people can be trusted I think that most people would try to be fair I would say that most of the time, most people try to be helpful Motivated Trust 6.7 6.8 6.9 It is beneficial for me to trust police and comply because they provide social order It is beneficial for me to trust the police and comply because they provide protection that can keep me safe It is beneficial for me to trust the police and comply because they provide civil services 6.10 It is beneficial for me to trust the police and comply because disobeying them can lead to negative consequences (arrested) 82 This section regards your perception of Confucian social values. Please select one between 1 ~ 7. No. Social Hierarchy (Deference to Authority) 7.1 Even if parent’s demands are Content unreasonable, children still should do what they ask 7.2 When elder and young come into a conflict, even if the elder is in the wrong, it is better for the young to just obey the elder 7.3 Being a student, one should not question the authority of their teacher Social Harmony (Conflict Aversion) 7.4 In a group, we should avoid open quarrel to preserve the harmony of the group 7.5 Even if there is some disagreement with others, it is better to avoid the conflict 7.6 A person should not insist on their own opinion if the group members already disagreed with it previously Group Primacy (Collective Welfare) 7.7 For the sake of the family, an individual should put his personal interests second 7.8 We should sacrifice our individual interests for the sake of the group’s collective interests 7.9 For the safety of the society, I am okay with sacrificing little liberty 1 2 3 4 5 6 7 83 This section regards your likelihood of compliance and cooperation with police. Please select one between 1~ 7. 1= Very Unlikely 2= Unlikely 3= Slightly Unlikely 4= Neutral 5= Slightly Likely 6= Likely 7= Very Likely No. Compliance with Law 8.1 How likely would you park your car Content illegally? 8.2 How likely would you dispose of your trash and litter illegally? 8.3 How likely would you speed or break traffic laws? 8.4 How likely would you not steal items from stores? 8.5 How likely would you download digital contents illegally? Compliance with Police 8.6 How likely would you follow the officer’s orders when they tell you to pull over? 8.7 How likely would you follow the officer’s order when they conduct Breathalyzer tests? (Refusing to blow hard enough) 8.8 How likely would you follow the officer’s orders when they demand identification during stops and questions? 8.9 How likely would you follow the officer’s orders when they order you to vacate the premise without explaining the reason? 8.10 How likely would you follow the officer’s orders when they ask you to go to the police station for questions? (not being arrested) Cooperation with Police 9.1 How likely would you call the police to report a crime? 9.2 How likely would you report a suspicious activity near your house to the police? 9.3 How likely would you call the police to report an accident? 9.4 How likely would you provide information to the police to help find a suspected criminal? 84 1 2 3 4 5 6 7 This section regards your background information. Please select among the options. Content No. Background Information 10.1 Current Academic Year (Check one) Freshmen Sophomore Junior Senior 1 2 3 4 Graduate School 5 10.2 Current Academic GPA (Check one) Below 1.0 1.0 ~ 2.0 2.0~ 3.0 3.0 ~ 4.0 Above 4.0 1 2 3 4 5 10.3 Parent’s Education (Check one) High School Associate Bachelor’s Degree Degree Master’s Degree Doctorate Degree 1 2 3 4 5 10.4 Household Annual Income (Parent’s Salary) Less than $15,000 $15,000 ~ $30,000 $30,000 ~ $45,000 $45,000 ~ $60,000 $60,000 ~ $75,000 Above $75,000 10.5 Gender 10.6 Age 1 2 3 4 5 6 Male 1 21 3 19 1 20 2 22 4 23 5 24 6 25 7 Female 2 26 8 27 9 28+ 10 1= Very Liberal 2= Liberal 3= Slightly Liberal 4= Moderate 5= Slightly Conservative 6= Conservative 7= Very Conservative Content No. Your Political Ideology 10.7 Where would you rank your political ideology regarding SOCIAL issues? 10.8 Where would you rank your political ideology regarding ECONOMIC issues? 10.9 Where would you rank your political ideology regarding GENERAL issues? Parent’s Political Ideology 10.10 Where would you rank your parent’s political ideology regarding SOCIAL issues? 10.11 Where would you rank your parent’s political ideology regarding ECONOMIC issues? 10.12 Where would you rank your parent’s political ideology regarding GENERAL issues? 85 1 2 3 4 5 6 7 APPENDIX B. Survey Instrument Korean Police Legitimacy Survey Korean (Dissertation) 설문에 참가해 주셔서 감사합니다. 이 설문의 목적은 경찰의 합법성에 대한 연구를 위한 것입니다. 연구자나 그 어느 누구도 참가자의 신원과 이름을 알 수 없습니다. 모든 참가자는 익명으로 처리 됩니다. 귀하의 프라이버시는 법으로 허용되는 한도 내에서 보호됩니다. 설문이 끝나면 꼭 반납해 주시기 바랍니다. 기밀유지 이 설문은 자발적인 설문이고 대략 20 분 정도 걸릴 것입니다. 만약에 하지 않더라도 그 어떤 패널티도 없을것이며 참가한다 하여 어떤 이득도 없을 것입니다. 설문을 익명으로 유지하기 위해 이름이나 학번 같은 정보를 공유하지 말아 주십시오. 완성된 설문은 데이터화 한 후, 설문과 데이터 모두 Michigan State University 오피스에 3 년간 보관 될 것이며 정보는 Institutional Review Board (IRB) 에게만 공유될 것입니다. 데이터화 된 데이터는 관리자 오피스에 잠금장치로 보안될 것입니다. 연구에대한 질문이나 걱정이 있으면, (Baker Hall, 655 Auditorium Road, Room 139, East Lansing, MI 48824, leesungu@msu.edu, 517-353-5150) 로 연락 주시기 바랍니다. 이 설문을 관리자에게 돌려줌으로서 자발적인 설문입을 동의 합니다. 86 이 설문은 본인의 경찰의 절차적 정의에 대한 의견입니다. 1 ~ 7 사이의 수 하나를 선택해 주십시오. 1= 강하게 동의하지 않음 2= 동의하지 않음 3= 조금 동의하지 않음 4= 중립 5= 조금 동의함 6= 동의함 7= 강하게 동의함 No. Content 1 2 3 4 5 6 7 절차적 정의 1.1 경찰은 시민을 존경으로 대한다 1.2 경찰은 모든 사람들 동등하게 대한다 1.3 경찰은 시민과 상호작용 할 때 정직하다 1.4 경찰은 시민의 요청이 있을 때는 경찰의 결정을 설명 한다 (왜 잡았는지 설명) 1.5 경찰은 시민이 의견을 발언할 수 있는 기회를 준다 1.6 경찰은 시민과 상호작용할 때는 중립적으로 행동한다 1.7 경찰은 자기들의 권한의 범위를 넘는다 1.8 경찰은 마치 자기들이 법 위에 있는 것처럼 행동한다 1.9 경찰은 정당한 이유도 없이 시민을 귀찮게 한다 87 이 부분은 경찰의 분배적 정의에 대한 설문입니다. 1 ~ 7 사이 수 하나를 골라주세요. No. Content 1 2 3 4 5 6 7 분배적 정의 2.1 경찰은 사람들을 대할 때 일관되게 법을 집행한다 2.2 경찰은 모든 시민에게 동질의 서비스를 제공한다 2.3 경찰은 시민들이 법률에 의하여 받아야 마땅한 결과물을 확실하게 받도록 노력한다 2.4 경찰은 정치인과 권력을 가진 사람들에게 더 많은 서비스를 제공한다 2.5 경찰은 부유한 시민들에게 더 좋은 서비스를 제공한다 이부분은 경찰의 업무수행 효율성에 대한 설문입니다. 1 ~7 사이 하나를 선택 해 주세요. No. Content 1 2 3 4 5 6 7 경찰 엄무수행 3.1 경찰은 우리 동네의 범죄를 효율적으로 다른다 3.2 경찰은 우리 동네의 폭력범죄를 효율적으로 다룬다 3.3 경찰은 우리 동네의 마약범죄를 효율적으로 다룬다 3.4 경찰은 우리 동네의 재산범죄를 효율적으로 다룬다 88 이 부분은 경찰의 믿을만한 가치와 규범적 노선에 대한 설문입니다. 1 ~ 7 하나를 골라 주세요. No. Content 1 2 3 4 5 6 7 믿을만한 가치 4.1 우리 동네의 경찰은 강력한 도덕강령을 가지고 엄무를 접한다 4.2 우리 동네의 경찰은 동네 주민들을 잘 보살 핀다 4.3 우리 동네의 경찰은 자신의 직무를 수행하는데 필요한 기술들을 가지고 있다 규범적 노선 4.4 경찰은 일반적으로 옳고 그름에 대하여 나와 동일한 감각을 가지고 있다 4.5 나와 경찰은 우리 동네에 대하여 바라는 바가 같다 4.6 우리 동네에 근무하는 대부분 경찰관들의 가치관은 나와 비슷하다 89 이 부분은 법률준수 의무와 경찰 복종의무에 대한 설문입니다. 1 ~ 7 중 하나를 골라주세요. No. Content 1 2 3 4 5 6 7 법률 준수의무 5.1 모든 법률은 엄격하게 준수되어야 한다 5.2 사람들은 법률이 요구한는 것을 행해야 한다 5.3 법을 어기는 사람은 사회의 다른 사람들에게 위험한 존재이다 5.4 법을 지키는 것은 궁극적으로 사회의 모든 사람들에게 이익이다 경찰 복종의무 5.5 귀하는 심지어 경찰에 동의하지 않더라도 경찰은 결정을 지지해야 한다 5.6 심지어 이유를 이해하지 못하거나 이유에 동의하지 못하더라도 경찰이 명령 하는것은 행해야 한다 5.7 비록 경찰이 귀하를 대하는 방식을 좋아하지 않더라도 경찰이 명령 하는것은 행해야 한다 5.8 귀하네 동네의 경찰은 정당한 권력이고 따라서 귀하는 경찰이 명령 하는것을 행해야 한다. 90 이 부분은 경찰 실뢰에 대한 설문 입니다. 1 ~ 7 중 하나를 골라주세요. No. Content 1 2 3 4 5 6 7 신뢰 6.1 나는 일반적으로 경찰의 판단에 의해 내가 약해질 수 있다는 것에 불편하지 않다 (경찰이 체포한다면 체포 당할 수 밖에 없는 나의 약함에 대하여 불편하지 않다) 6.2 나에게 중요한 특수한 상황들을 경찰이 다루는 것에 대해 불편하지 않다 6.3 나는 질서유지를 위한 경찰의 의사결정들에 대해서 불편해 하지 않는다 신뢰 성향 6.4 대체로 말해서 나는 대부분의 사람들은 믿을 수 있다고 생각한다 6.5 나는 대부분의 사람들이 공정하려고 노력한다고 생각한다 91 6.6 나는 대부분의 사람들은 대부분의 경우 도움이 되려고 노력한다고 생각한다 이득에 의한 신뢰 6.7 경찰은 사회 질서유지를 하기에 경찰의 믿고 따르는게 나에게 더 이득이다 6.8 경찰은 나를 보호해 주고 안전하게 해 줌으로, 경찰을 믿고 따르는것이 나에게 더 이득이다 6.9 경찰은 많은 서비스를 제공해 줌으로 경찰을 믿고 따르는 것이 나에게 더 이득이다 6.10 경찰의 말을 복종하지 않으면 체포와 같이 불이득을 당할수 있기에 경찰의 말들 믿고 따르는게 나에겐 더 이득이다 이 부분은 유교사상에 대한 설문입니다. 1 ~ 7 중 하나를 골라주세요. No. Content 1 2 3 4 5 6 7 어른 공경 7.1 부모님의 요구가 이해가 안되더라도, 자식의 도리로서 따라야 한다 7.2 어른과 아이가 다툰다면, 어린이 틀렸더라도 어른의 말을 듣는게 낫다 7.3 학생으로서 선생님이나 교수님의 권위에 도전해서는 안된다 사회적 조화 7.4 단체 생활에서, 집단의 조화를 위하여 싸우는 것은 좋지 않다 7.5 집단생활 하는데 남의 의견과 동의하지 않더라도 다툼을 피하는 것이 낫다 92 7.6 단체생활시 타인의 의견도 존중해야지 내의견만 내새우려고 하면 안된다 집단적 중요성 7.7 가족을 위해서라면 개인적 이윤을 뒤로해도 된다 7.8 집단의 더 큰 이윤을 위해서라면 나의 개인적인 이윤은 희생할수 있다 7.9 사회의 안전을 위해서라면, 나의 자유를 조금 희생해도 괜찬다 이 부분은 법률준수, 경찰에 순응과 협조에 관한 설문입니다. 1 ~ 7 중 골라주세요. 1= 매우 그렇지 않을 것이다 2= 그렇지 않을 것이다 3= 조금 그렇지 않을 것이다 4= 중립 5= 조금 그럴 것이다 6= 그럴 것이다 7= 매우 그럴 것이다 No. Content 1 2 3 4 5 6 7 법률 준수 8.1 자동차를 불법 주차를 할 확률이 얼마나 되십니까? 8.2 쓰래기를 불법 투기 할 확률이 얼마나 되십니까? 8.3 교통 법규를 어기거나 과속을 할 확률이 얼마나 되십니까? 93 8.4 가게에서 물건을 훔칠 확률이 얼마나 되십니까? 8.5 디치털 콘텐츠를 불법으로 내려 받을 확률이 얼마나 되십니까? 경찰에 대한 순응 8.6 경찰이 나에게 자도앛를 갓길에 멈추라고 할 때 명령을 따를 확률이 얼마나 되십니까? 8.7 경찰이 음주측정을 할 때 명령을 따를 확률이 얼마나 되십니까? (예 충분히 쎄게 불지 않는 등 시간 끌기 등) 8.8 경찰이 불심검문 시 신분증을 요구할 때 명령을 따를 확률이 얼마나 되십니까? 8.9 경찰이 이유도 설명하지 않고 현장을 비우라고 명령할 때 그 명령을 따를 확률이 얼마나 되십니까? 8.10 체포되지 않았는데 심문하려고 경찰서로 동행을 요구할 때 그 명령을 따를 확률이 얼마나 되십니까? 경찰에 협조 9.1 범죄를 신고하기 위하여 경찰에 전화를 할 확률은 얼마나 되십니까? 9.2 집 근처에서 일어나고 있는 의심스러운 활동을 경찰에 신고할 확률이 얼마나 되십니까? 9.3 사고를 신고하기 위하여 경찰에 전화를 할 확률이 얼마나 되십니까? 9.4 범죄 용의자를 찾는데 도움이 되는 정보를 경찰에 제보할 확률이 얼마나 되십니까? 94 이 부분은 배경정보를 위한 설문 입니다. 1 ~ 7 중 하나를 골라주세요. No. Content 배경정보 10.1 현재 학년 1 학년 2 학년 3 학년 4 학년 대학원 1 2 3 4 5 10.2 현재 성적 1.0 이하 1.0 ~ 2.0 2.0~ 3.0 3.0 ~ 4.0 1 2 3 4 4.0 이상 5 10.3 부모님 고졸 전분대 졸 4 년제 졸 교육정도 1 2 3 석사 4 박사 5 10.4 부모님 1500 만원 연봉 이하 10.5 성별 10.6 나이 1 19 1 20 2 1500 만원 ~ 3000 만원 3000 만원 ~ 4500 만원 4500 만원 ~ 6000 만원 6000 만원 ~ 7500 만원 7500 만원 이상 2 3 4 5 6 남자 1 21 3 22 4 23 5 24 6 25 7 여자 2 26 8 27 9 28+ 10 1= 매우 진보 2= 진보 3= 조금 진보 4= 중간 5= 조금 보수 6= 보수 7= 매우 보수 No. Content 1 2 3 4 5 6 7 정치이념 10.7 사회적 문제를 보았을때 본인의 정치 이념은 어떻습니까? 10.8 경제적 문제를 보았을때 본인의 정치 이념은 어떻습니까? 10.9 전반적인 문제를 보았을때 본인의 정치 이념은 어떻습니까? 부모님 정치 이념 10.10 사회적 문제를 보았을때 부모님의 정치 이념은 어떻습니까? 95 10.11 경제적 문제를 보았을때 부모님의 정치 이념은 어떻습니까? 10.12 전반적인 문제를 보았을때 부모님의 정치 이념은 어떻습니까? 96 REFERENCES 97 REFERENCES Aberbach, J., & Walker, J. (1970). Political trust and racial ideology. The American Political Science Review, 64(4), 1199-1219. Alarcon, G., Lyons, J., & Christensen, J. (2016). The effect of propensity to trust and familiarity on perceptions of trustworthiness over time. Personality and Individual Differences, 94, 309-315. Antrobus, E., Bradford, B., Murphy, K., & Sargeant, E. (2015). Community norms, procedural justice, and the public's perceptions of police legitimacy. Journal of Contemporary Criminal Justice, 31(2), 151-170. Baker, T., & Gau, J. (2018). Female offenders' perceptions of police procedural justice and their obligation to obey the law. Crime & Delinquency, 64(6), 758-781. Barkworth, J., & Murphy, K. (2015). Procedural justice policing and citizen compliance behavior: The importance of emotion. Psychology, Crime & Law, 21(3), 254-273. Beetham, D. (1991). The legitimation of power. London, England: Macmillan. Berg, J., Dickhaut, J., & McCabe, K. (1995). Trust, reciprocity, and social history. Games and Economic Behavior, 10(1), 122-142. Blader, S., & Tyler, T. (2003). A four component model of procedural Justice: Defining the meaning of a fair process. Personality & Social Psyhology Bulletin, 29(36), 747-758. Bottoms, A., & Tankebe, J. (2012). Beyond procedural justice: A diologic approach to legitimacy in criminal justice. The Journal of Criminal Law & Criminology, 102, 119- 170. Butler, J. (1999). Trust expectations, information sharing, climate of trust and negotiation effectiveness and efficacy. Group & Organization Management, 24, 217-238. Butler, G., & Mathews, A. (1983) Cognitive processes in anxiety. Advances in Behavior Research and Therapy, 5, 51-62. Chao, Y. (1994). Culture and work organization: The Cinese Case. In S. Kao, D. Sinha, & S. Ng, Effective Organizations and Social Values (pp. 28-36). New Delhi: Sage. Chatman, J. (1991). Matching people and organizations: Seletion and socialization in public accounting firms. Administrative Science Quarterly, 36, 459-484. Chung, J., & Pysarchik, D. (2000). A model of behavioral intention to buy domestic versus imported products in a Confucian culture. Marketing Intelligence & Planning, 18(5), 281-291. Chung, K., Eichenseher, J., & Taniquchi, T. (2008). Ethical perceptions of business students: Differences between East Asia and the USA and among "Confucian" cultures. Journal of Business Ethics, 79, 121-132. 98 Colquitt, J., Scott, B., & LePine, J. (2007). Trust, trustworthiness, and trust propensity: A metal- analytic test of their unique relationships with risk taking and job performance. Journal of Applied Psychology, 98, 909-927. Costa, T., & McCrae, R. (1992). Revised NEO Personality Inventory and NEO Five-factor Inventory Professional Manual. Odesa: Psychological Assessment Resources. Dai, M., Frank, J., & Sun, I. (2011). Procedural justice during police-citizen encounters: The effects of process-based policing on citizen compliance and demeanor. Journal of Criminal Justice, 39(2), 159-168. Dalton, R., & Ong, n. (2005). Authority orientations and democratic attitudes: A test of the 'Asian Values' hypothesis. Japanese Journal of Political Science, 6(2), 1-21. De Cremer, D., & Tyler, T. (2007). The effects of trust in authority and procedural fairness on cooperation. Journal of Applied Psychology, 92(3), 639-649. Fagan, J., & Tyler, T. R. (2005). Legal socialization of children and adolescents. Social justice research, 18(3), 217-241. Frazier, L., Johnson, P., & Fainshmidt, S. (2013). Development and validation of a propensity to trust scale. Journal of Trust Research, 3(2), 76-97. Gau, J. (2011). The convergent and discriminant validity of procedural justice and police legitimacy: An empirical test of core theoretical propositions. Journal of Criminal Justice, 39(6), 489-498. Gau, J. (2013). Consent searches as a threat to procedural justice and police legitimacy: An analysis of consent requests during traffic stops. Criminal Justice Policy Review, 24(6), 759-777. Gau, J. (2014). Procedural justice and police legitimacy: A test of measurement and structure. American Journal of Criminal Justice, 39, 187-205. Gau, J. (2015). Procedural justice, police legitimacy, and legal cynicism: A test for mediation effects. Police Practice and Research: An International Journal, 16(5), 402-415. Gau, J., & Brunson, R. (2010). Procedural justice and order maintenance policing: A study of inner-city young men's perceptions of police legitimacy. Justice Quarterly, 27(2), 255- 279. Gau, J., Corsaro, N., Stewart, E., & Brunson, R. (2012). Examining macro-level impacts on procedural justice and police legitimacy. Journal of Criminal Justice, 40(4), 333-343. Gefen, D., & Heart, T. (2006). On the need to include national cultures as a central issue in e- commerce trust beliefs. Journal of Global Information Management, 14(4), 1-30. Gerber, M., & Jackson, J. (2016). Justifying violence: Legitimacy, ideology and public support for police use of force. Psychology, Crime & Law, 23(1), 79-95. Gravetter, F. J., & Wallnau, L. B. (2014). Essentials of statistics for the behavioral sciences (8th ed ed.). Belmont, CA: Wadsworth. 99 Greenberg, R., Wong, B., & Lui, G. (2008). Culture and consumer trust in online business. Journal of Global Information Management, 16(3), 26-44. Hamm, J., PytlikZillig, L., Herian, M., Bornstein, B., Tomkins, A., & Hoffman, L. (2013). Deconstructing public confidence in state courts. Journal of Trust Research, 3(1), 11-31. Hamm, J., Trinkner, R., & Carr, J. (2017). Fair process, trust, and cooperation: Moving toward an integrated framework of police legitimacy. Criminal Justice and Behavior, 44(9), 1183-1212. Hanson, D. W., & Fowler, R. B. (Eds.). (1971). Obligation and Dissent: An Introduction to Politics. Little, Brown. Hilbig, B., & Zettler, I. (2009). Pillars of cooperation: Honesty-humility, social value orientations, and economic behavior. Journal of Research in Personality, 43, 516-519. Hilbig, E., Zettler, I., Leist, F., & Hedasch, T. (2013). It takes two: Honesty-humility and agreeableness differentially predict active versus reactive cooperation. Personality and Individual Differences, 54, 598-603. Hinds, L., & Murphy, K. (2007). Public satisfaction with police: Using procedural justice to improve police legitimacy. Australian & New Zealand Journal of Criminology, 40(1), 27- 42. Hinds, L. (2009). Youth, police legitimacy and informal contact. Journal of Police and Criminal Psychology, 24, 10-21. Hough, M., Jackson, J., Bradford, B., Myhill, A., & Quinton, P. (2010). Procedural justice, trust and institutional legitimacy. Policing: A Journal of Policy and Practice, 4(3), 203-210. Hu, L. T., & Bentler, P. M. (1999). Cutoff criteria for fit indexes in covariance structure analysis: Conventional criteria versus new alternatives. Structural Equation Modeling: A Multidisciplinary Journal, 6(1), 1-55. Huq, A., Jackson, J., & Trinkner, R. (2016). Legitimating practicies: Revisiting the predicates of police legitimacy. The British Journal of Criminology, 5(1), 1101-1122. Hyun, K. (2001). Sociocultural change and traditional values: Confucian values among Koreans and Korean Americans. International Journal of Intercultural Relations, 25,(2), 203-229. Jackson, J., Bradford, B., Hough, M., Myhill, A., Quinton, P., & Tyler, T. (2012). Why do people comply with the law? Legitimacy and the influence of legal institutions. The British Journal of Criminology, 52(6), 1051-1071. Jackson, J., & Gau, J. (2016). Carving up concepts? Differentiating between trust and legitimacy in public attitudes towards legal authority. In E. Shockley, T. Neal, L. PytlikZillig, & B. Bornstein, Interdisciplinary Perspectives on Trust: Towards Theoretical and Methodological Integration. New York: Springer. Jackson, J., Huq, A., Bradford, B., & Tyler, T. (2013). Monopolizing force? Police legitimacy and public attitudes toward the acceptability of violence. Psychology, Public Policy and Law, 19(4), 479-497. 100 Jackson, J., Milani, J., & Bradford, B. (2018). Empirical legitimacy and normative compliance with the law. Global encyclopedia of public administration, public policy, and governance. Jarvenpaa, S., & Tractinsky, N. (2003). Consumer trust in an internet store: A cross-cultural validation. In C. Steinfeld, New directions in research on e-commerce (pp. 33-63). West Lafayette, IN: Purdue University Press. Jin, J., & Kim, D. (2018). Expansion of the subway network and spatial distribution of population and employment in the Seoul metropolitan area. Urban Studies, 55(11), 2499- 2521. Jonathan-Zamir, T., & Weisburd, D. (2013). The effects of security threats on antecedents of police legitimacy: Findings from a quasi-experiment in Israel. 50(1), 3-32. Keller, G., & Kronstedt, C. (2005). Connecting Confucianism, communism and the Chinese culture of commerce. The Journal of Language for International usiness, 16(1), 60-75. Kenneth, T. (2019, 4 02). WorldAtlas. Retrieved from Ethnic Minorities and Immigrants in South Korea: https://www.worldatlas.com/articles/ethnic-minorities-and-immigrants-in-south- korea.html Kim, B., Li, L., & NG, G. (2005). The Asian American values scale-Multidimensional: Development, Reliability, and Validity. Cultural Diversity and Ethnic Minority Psychology, 11(3), 187-201. Kim, Y., Ra, K., & McLean, K. (2018). The generalizability of police legitimacy: Procedural justice, legitimacy, and speeding intention of South Korean drivers. Asian Journal of Criminology, 14, 41-59. Klepper, S., & Nagin, D. (1989). The deterrent effect of perceived certainty and severity of punishment revisited. Criminology, 27(4), 721-746. Kochel, T. (2014). Causes of police legitimacy. In G. Bruinsma, & D. Weisburd, Encyclopedia of criminology and criminal justice (pp. 295-303). New York, NY: Springer. Kochel, T., Parks, R., & Mastrofski, S. (2013). Examining police effectiveness as a precursor to legitimacy and cooperation with police. Justice Quarterly, 30(5), 895-925. Korean Educational Development Institute (2019). Higher Educaiton Statistics Survey. Seoul, South Korea. Retrieved from https://hi.kedi.re.kr/home/univ/school2. Krueger, J., Massey, A., & DiDonato, T. (2008). A matter of trust: From social preferences to the strategic adherence to social norms. Negotiation and Conflict Management Research, 1(1), 31-52. Lawson, C., & MacLeod, C. (1999). Depression and the interpretation of ambiguity. Behavior Research and Theraphy, 37, 463-474. Lee, Y., & Cho, S. (2019). The significance of instrumental pathways to legitimacy and public support for policing in South Korea: Is the role of procedural fairness too small? Crime, Law and Social CHange. doi:https://doi.org/10.1007/s10611-019-09876-z 101 Lebel, E., & Paunonen, S. (2011). Sexy but often unreliable: The impact of unreliability on the replicability of experimental findings with implicit measures. Personality and Social Psychology Bulletin, 37(4), 570-583. Levy, M. (1992). Confucianism and Modernization. Society, 29(4), 15-18. Levi, M., Sacks, A., & Tyler, T. (2009). Conceptualizing legitimacy, measuring legitimating beliefs. American Behavioral Scientist, 53(3), 354-375. MacCoun, R. (1993). Drugs and the Law: A psychological analysis of drug prohibition. Psychological Bulletin, 113, 497-512. Madon, N. S., Murphy, K., & Cherney, A. (2017). Promoting community collaboration in counterterrorism: Do social identities and perceptions of legitimacy mediate reactions to procedural justice policing?. British journal of criminology, 57(5), 1144-1164. Mani, D. (2018, October 18). World Education News and Reviews. Retrieved from Education in South Korea: https://wenr.wes.org/2018/10/education-in-south-korea Mayer, R., & Davis, J. (1999). The effects of the performance appraisal system on trust for management: A field quasi-experiment. Journal of Applied Psychology, 84, 123-136. Mayer, R., Davis, J., & Schoorman, F. (1995). An integrative model of organizational trust. The Academy of Management Review, 20, 709-734. Mazerolle, L., Antrobus, E., Bennett, S., & Tyler, T. (2013). Shaping citizen perceptions of police legitimacy: A randomized field trial of procedural justice. Criminology, 51(1), 33- 63. Mazerolle, L., Bennett, S., Antrobus, E., & Eggins, E. (2012). Procedural justice, routine encounters and citizen perceptions of police: Main findings from Queensland Community Engagement Trial. Journal of Experimental Criminology, 8(4), 343-367. Mazerolle, L., Sargeant, E., Cherney, A., Bennett, S., Murphy, K., Antrobus, E., & Martin, P. (2014). Why police should care about procedural justice and legitimacy. In L. Mazerolle, E. Sargeant, A. Cherney, S. Bennett, K. Murphy, E. Antrobus, & P. Martin, Procedural Justice and Legitimacy in Policing (pp. 1-11). Cham, Switzerland: Springer International Publishing. Mogg, K., Bradbury, K., & Bradley, B. (2006). Interpretation of ambiguous information in clinical depression. Behavior Research and Theraphy, 44, 1411-1419. Monkhouse, L., Barnes, B., & Pham, T. (2013). Measuring Confucian values among East Asian consumers: a four country study. Asia Pacific Business Review, 19(3), 320-336. Mooradian, T., Renzl, B., & Matzler, K. (2006). Who trusts? Personality, trust and knowledge sharing. Management Learning, 37(4), 523-540. Murphy, K. (2005). Regulating more effectively: The relationship between procedural justice, legitimacy, and tax non-compliance. Journal of Law and Society, 32(4), 562-589. Murphy, K. (2013). Policing at the margins: Fostering trust and cooperation among ethnic minority groups. Journal of Policing, Intelligence and Counter Terrorism, 8(2), 184-199. 102 Murphy, K., & Cherney, A. (2012). Understanding cooperation with police in a diverse society. The British Journal of Criminology, 52(1), 181-201. Murphy, K., Hinds, L., & Fleming, J. (2008). Encouraging public cooperation and support for police. Policing & Society, 18(2), 136-155. Murphy, K., & Tyler, T. (2008). Procedural justice and compliance behavior: The mediating role of emotions. European Journal of Social Psychology, 38(4), 652-668. Murphy, K., Mazerolle, L., & Bennertt, S. (2014). Promoting trust in police: findings from a randomized experimental field trail of procedural justice policing. Policing and Society, 24(4), 405-424. Murphy, K., Sargeant, E., & Cherney, A. (2015). The importance of procedural justice and police performance in shaping intentions to cooperate with the police: Does social identity matter? European Journal of Criminology, 12(6), 719-738. Murphy, K., Tyler, T., & Curtis, A. (2009). Nurturing regulatory compliance: Is procedural justice effective when people question the legitimacy of the law? Regulation & Governance, 3(1), 1-26. Murphy, K., Bradford, B., & Jackson, J. (2016). Motivating compliance behavior among tax offenders: Procedural justice or deterrence? Criminal Justice and Behavior, 43(1), 102- 118. Na, E., & Cha, J. (1999). Changes in values and the generation gap between the 1970s and the 1990s in Korea. Korean Journal of Psychology: Social Psychology and Personality, 13(2), 37-60. Nagin, D. (1988). Criminal deterrence research at the outset of the twenty-first century. Crime & Justice, 23, 1-42. Nagin, D., & Paternoster, R. (1991). On the relationship of past to future participation in delinquency. Criminology, 29(2), 163-189. Nagin, D., & Pogarsky, G. (2003). An experimental investigation of deterrence: Cheating, self- serving bias, and impulsivity. Criminology, 41(1), 167-194. Nagin, D., & Telep, C. (2017). Procedural justice and legal compliance. Annual Review of Law and Social Science, 13, 5-28. Nambudiri, R. (2012). Propensity to trust and organizational commitment: A study in the Indian pharmaceutical sector. The International Journal of Human Resource Management, 23(5), 977-986. Oetzel, J., & Ting-Toomy, S. (2003). Face concerns in interpersonal conflict: A Cross-cultural empirical test of the Face Negotiation Theory. Communications Research, 30(6), 599- 624. Paek, S., Nalla, M., & Lee, J. (2018). Perception of police legitimacy among private security officers. Security Journal, 1-19. 103 Paige, J. (1971). Political orientation and riot participation. American Sociological Review, 36(5), 810-820. Park, M., & Chesla, C. (2007). Revisiting Confucianism as a conceptual framework for Asian family study. Journal of Family Nursing, 13(3), 293-311. Park, C., & Shine, D. (2006). Do Asian values deter popular support for democracy in South Korea? Asian Survey, 46(3), 341-361. Paternoster, R. (1987). The deterrent effect of the perceived certainty and severity of punishment: A review of evidence and issues. Justice Quarterly, 4(2), 173-217. Paternoster, R. (1989). Decisions to participate in and desist from four types of common delinquency: Deterrence and rational choice perspective. Law & Society Review, 23(1), 7- 40. Paternoster, R., & Iovanni, L. (1986). The deterrent effect of perceived severity: A re- examination. Social Forces, 64(3), 751-777. Paternoster, R., Saltzman, L., Chiricos, T., & Waldo, G. (1982). Perceived risk and deterrence: Methodological artifacts in perceptual deterrence research. Journal of Criminal Law & Criminology, 73(3), 1238-1258. Paternoster, R., Saltzman, L., Waldo, G., & Chiricos, T. (1983). Perceived risk and social control: Do sanctions really matter? Law & Society Review, 17(3), 457-479. Penner, E., Viljoen, J., Douglas, K., & Roesch, R. (2014). Procedural justice versus risk factors for offending: Predicting recidivism in youth. Law and Human Behavior, 44(5), 733-755. Phuong-Mai, N., Terlouw, C., & Pilot, A. (2005). Cooperative learning vs Confucian heritage culture's collectivism: Confrontation to reveal some cultural conflicts and mismatch. American Economic Journal, 3, 403-419. Pickett, J., & Bontrager, S. (2017). Procedurally just cooperation: Explaining support for due process reforms in policing. Journal of Criminal Justice, 48(1), 9-20. Pytlikzillig, L., Hamm, J., Shockley, E., Herian, M., Neal, T., Kimbrough, C., . . . Bornstein, B. (2016). The dimensionality of trust-relevant constructs in four institutional domains: Results from confirmatory factor analyses. Journal of Trust Research, 6, 111-150. Reisig, M., & Lloyd, C. (2009). Procedural justice, police legitimacy, and helping the police fight crime: Results fromm a survey of Jamaican adolescents. Police Quarterly, 12(1), 42-62. Reisig, M., Bratton, J., & Gertz, M. (2007). The construct validity and refinement of process- based policing measures. Criminal Justice and Behavior, 34, 1005-1028. Reisig, M., Tankebe, J., & Mesko, G. (2014). Compliance with the law in Slovenia: The role of procedural justice and police legitimacy. European Journal of Criminal Justice Policy Research, 20, 259-276. 104 Reisig, M., Wolfe, S., & Holtfreter, K. (2012). Legal cynicism, legitimacy, and criminal offending: The non-confounding effect of low self-control. Criminal Justice & Behavior, 38, 1170-1184. Rosenbaum, D., Lawrence, D., Hartnett, S., McDevitt, J., & Posick, C. (2015). Measuring procedural justice and legitimacy at the local level: The police-community interaction survey. Journal of Experimental Criminology, 11(3), 335-366. Rotter, J. (1971). Generalized expectations for interpersonal trust. American Psychologist, 26, 443-452. Rotter, J. (1980). Interpersonal trust, trustworthiness, and gullibility. American Psychologist , 35, 1-7. Rowatt, W., Powers, C., Targhetta, V., Comer, J., Kenneydy, S., & Labouff, J. (2006). Development and initial validations of an implicit measure of humility relative to arrogance. The Journal of Positive Psychology, 1, 198-211. Rowley, C., & Warner, M.(2010). Managers, markets and the globalized economy in China: Epilogue. Asia Pacific Business Review, 16(3), 483-491. Sampson, Raudenbush, & Earl. (1997). Neighborhoods and violent crime: A multi-level study of collective efficacy. Science, 277, 918-924. Sarat, A. (1977). Studying American legal culture. Law & Society, 11, 427-488. Sargeant, E., Wickes, R., & Mazerolle, L. (2013). Policing community problems: Exploring the role of formal social control in shaping collective efficacy . Australian & New Zealand Journal of Criminology, 46(1), 70-87. Seo, B., & Lee, S. (2018). Mediating effect of cooperation with the police on compliance with the law in model of procedural justice and police legitimacy. International Journal of Advanced Culture Technology, 6(1), 15-22. Sherman, L. (1993). Defiance, deterrence, and irrelevance: A theory of the criminal sanction. Journal of Research in Crime and Delinquency, 30(4), 445-473. Shin, G., & Moon, R. (2019). Korea's migrants: From homegeneity to diversity. Asian Survey, 59(4), 595-606. Shin, C., & Shaw, I. (2003). Social policy in South Korea: Cultural and structural factors in the emergence of welfare. Social Policy & Administration, 37(4), 328-341. Soper, D. (2019, 12 15). A-priori Sample Szie Calculator for Structural Equation Models. Retrieved from http://www.danielsoper.com/statcalc Suh, M. (2019). Two sacred tales in the Seoul metropolis: The gospels of prosperity and development in modernizing South Korea. Social Compass, 66(4), 561-578. Sunshine, J., & Tyler, T. (2003). The role of procedural justice and legitimacy in shaping public support for policing. Law and Society Review, 37, 513-548. 105 Tan, F., & Sutherland, P. (2004). Online consumer trust: A multi-dimensional model. Journal of Electronic Commerce in Organizations, 2(3), 40-58. Tan, H., & Chee, D. (2005). Understanding interpersonal trust in confucian-influenced society. International Journal of Cross Cultural Management, 5(2), 197-212. Tankebe, J. (2009). Public cooperation with the police in Ghana: Does procedural fairness matter? Criminology, 47(4), 1265-1293. Tankebe, J. (2013. Viewing things differently: The dimensions of public perceptions of police legitimacy, Criminology, 51(1), 103-135. Tankebe, J. (2013). Viewing things differently: The dimensions of public perceptions of police legitimacy. Criminology, 51(1), 103-135. Tankebe, J., Reisig, M., & Wang, X. (2016). A multidimensional model of police legitimacy: A cross-cultural assessment. Law and Human Behavior, 40(1), 11-22. Thibaut, J., & Walker, L. (1975). Procedural Justice: A psychological analysis. Hillsdale, NJ: Erlbaum. Thielmann, I., & Hilbig, B. (2014). Trust in me, trust in you: A social projection account of the link between personality, cooperativeness, and trustorthiness expectations. Journal of Research in Personality, 50, 61-65. Ting-Toomy, S., Gao, G., Trubisky, P., Yang, Z., Kim, H., Lin, S.-L., & Nishida, T. (1991). Culture, Face Maintenance, and styles of handling interpersonal conflict: A study in five cultures. The International Journal of Conflict Management, 2(4), 275-296. Trinkner, R., & Tyler, T. (2018). Legal socialization: Coercion versus consent in an era of mistrust. Annual Review of Law and SOcial Science, 12, 417-439. Tsushima, M., & Hamai, K. (2015). Public cooperation with the police in Japan: Testing the legitimacy model. Journal of Contemporary Criminal Justice, 31(2), 212-228. Tu, W. (1998). Confucius and Confucianism. In W. Slote, & G. DeVos, Confucian and the family. New York: University of New York Press. Tyler. (1990). Why People Obey the Law. New Haven: Yale University Press. Tyler, T. (1988). What is procedural justice? Criteria used by citizens to assess the fairness of legal procedures. Law & Society Review, 22, 103-135. Tyler, T. (2001). Public trust and confidence in legal authorities: What do majority and minority group members want from the law and legal institutions? Behavioral Sciences and the Law, 19, 215-235. Tyler, T. (2003). Procedural justice, legitimacy, and the effective rule of law. Crime and Justice , 30, 283-357. Tyler, T. (2004). Enhancing police legitimacy. American Academy of Political and Social Sciences, 593, 84-99. 106 Tyler, T. (2005). Policing in black and white: Ethnic group differences in trust and confidence in the police. Police Quarterly, 8, 322-342. Tyler, T. (2006). Why people obey the law. Princeton: Princeton University Press. Tyler, T., & Blader, S. (2000). Cooperation in groups: Procedural justice, social identity, and behavioral engagement. Philadelphia, PA: Psychology Press. Tyler, T., & Huo, Y. (2002). Trust in the law: Encouraging public cooperation with the police and courts. New York: Russell Sage. Tyler, T., & Fagan, J. (2008). Legitimacy and Cooperation: Why do people help the police fight crime in their communities? Ohio State Journal of Criminal Law, 6, 231-275. Tyler, T., & Jackson, J. (2013). Future challenges in the study of legitimacy and criminal justice. Yale Law School, Public Law Working Paper, (264). Tyler, T., & Jackson, J. (2014). Popular legitimacy and the exercise of legal authority: Motivating compliance, cooperation, and engagement. Psychology, Public Policy, and Law, 20(1), 78-95. Tyler, T., & Wakslak, C. (2004). Profiling and police legitimacy: Procedural justice, attributions of motive, and acceptance of police authority. Criminology, 42(2), 253-280. Van Dyne, L., Vandewalle, D., Kostova, T., Latham, M., & Cummings, L. (2000). Collectivism, propensity to trust and self-esteem as predictors of organizational citizenship in a non- work setting. Journal of Organizational Behavior, 21, 3-23. Weber, M. (1968). Economy and society. In G. Roth, & C. Wittich, Economy and society: An outline of interpretive sociology. New York: Bedminster. Wee, C. (2014). Filial Obligations: A Comparative Study. Dao: A Journal of Comparative Philosophy, 13, 83-97. Weidman, A., Cheng, J., & Tracy, J. (2018). The psychological structure of humility. Journal of Personality and Social Psychology. 114(1), 153-178. Westland, J. (2010). Lower bounds on sample size in structural equation modeling. Electronic Commerce Research and Applications, 9(6), 476-487. Wilson, J., & Kelling, G. (1982). Broken windows. Atlantic Monthly, 249(3), 29-38. Wolfe, S. (2011). The effect of low self-control on perceived police legitimacy. Journal of Criminal Justice, 39, 67-74. Wolfe, S., McLean, K., & Pratt, T. (2016). I learned it by watching you: Legal socialization and the intergenerational transmission of legitimacy attitudes. British Journal of Criminology, 1-35. Woo, Y., Maguire, E., & Gau, J. (2018). Direct and indirect effects of procedural justice on cooperation and compliance: Evidence from South Korea. Police Practice and Research, 19(2), 168-185. Wright, J. (1976). The Dissent of the Governed. New York: Academic Press. 107 Yau, O. H. (1988). Chinese cultural values: Their dimensions and marketing implications. European Journal of marketing, 22(5), 44-57. Yim, H. (2002). Cultural identity and cultural policy in South Korea. The International Journal of Cultural Policy, 8(1), 37-48. Young, A., Lei, L., Wong, B., & Kwok, B. (2016). Individual tax compliance in China: A review. International Journal of Law and Management, 58(5), 562-574. Zettler, I., Hilbig, B., & Heydasch, T. (2013). Two sides of one coin: Honesty-humility and situational factors mutually shape social dilemma decision making. Journal of Research in Personality, 47, 286-295. Zhai, Y. (2017). Values of deference to authority in Japan and China. International Journal of Comparative Sociology, 58(2), 120-139 108