Open Collections

UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Grievance initiation in an ethnically diverse organization Au, Kevin Y. 1996

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Item Metadata

Download

Media
831-ubc_1996-090396.pdf [ 7.69MB ]
Metadata
JSON: 831-1.0087818.json
JSON-LD: 831-1.0087818-ld.json
RDF/XML (Pretty): 831-1.0087818-rdf.xml
RDF/JSON: 831-1.0087818-rdf.json
Turtle: 831-1.0087818-turtle.txt
N-Triples: 831-1.0087818-rdf-ntriples.txt
Original Record: 831-1.0087818-source.json
Full Text
831-1.0087818-fulltext.txt
Citation
831-1.0087818.ris

Full Text

GRIEVANCE INITIATION IN A N ETHNICALLY DIVERSE ORGANIZATION by KEVIN Y. A U B.B.A., The Chinese University of Hong Kong, 1989 M.Phil., The Chinese University of Hong Kong, 1991 A THESIS SUBMITTED IN PARTIAL FULFILMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY in THE FACULTY OF GRADUATE STUDIES Faculty of COMMERCE AND BUSINESS ADMINISTRATION We accept this thesis as conforming to the required standard THE UNIVERSITY OF BRITISH COLUMBIA December 1995 ® Kevin Y. Au, 1995 In presenting this thesis in partial fulfilment of the requirements for an advanced degree at the University of British Columbia, I agree that the Library shall make it freely available for reference and study. I further agree that permission for extensive copying of this thesis for scholarly purposes may be granted by the head of my department or by his or her representatives. It is understood that copying or publication of this thesis for financial gain shall not be allowed without my written permission. X2e|Sai«ttS$ of COMMERCE AND BUSINESS ADMINISTRATION The University of British Columbia Vancouver, Canada „ . DECEMBER 2 0 , 1 9 9 5 . DE-6 (2/88) Abstract The demographic composition of the North American workforce is changing due to the increase in non-European immigrants and different reproductive rates among ethnic groups. This trend has created the need for managers and academics to rethink traditional ways of management and derive new ways to integrate minority workers for better performance. Guided by a social psychological framework of multiculturalism, this project examined the use of the grievance procedure in an ethnically diverse organization. It tested various hypotheses concerning how ethnic belonging, perceived atmosphere of multiculturalism, and individualism-collectivism value affected job satisfaction, complaint behaviours of workers, the number of grievances filed, and workers' evaluations of the grievance procedure. A study was conducted in the Vancouver plant of a Crown Corporation. Through the cooperation of management and the union, labour officers, shop stewards, and supervisors of the plant were first interviewed. After that, 650 questionnaires were distributed to in-house, blue-collar workers who voluntarily participated in the study. Of the 139 questionnaires returned (21%), 130 of them were completed and analyzed by regression. Many of the hypotheses are supported. English ability is positively associated with ethnic belonging for majority workers, but not for minority workers (Linguistic Vitality Hypothesis). It is also positively associated with complaints to supervisors and the number of grievances filed, but not with complaints to coworkers of same ethnic background (Linguistic Accommodation Hypothesis). For minority workers, a stronger ethnic belonging is related to perceiving a positive atmosphere of multiculturalism (Ethnic Confidence Hypothesis). This relationship, however, does not exist for majority workers. Moreover, Ethnic Status ii Hypothesis is also supported, as ethnic status is positively associated with perceiving a more positive atmosphere of multiculturalism. Contrary to the general beliefs concerning multiculturalism, perceived atmosphere of multiculturalism has little direct effect on grievance behaviours in a multicultural workplace (Primary Multiculturalism Hypothesis). However, in line with the literature of situational ethnicity, the interaction of perceived atmosphere of multiculturalism and ethnic belonging exerts significant effects on grievance behaviours (Secondary Multiculturalism Hypothesis). In particular, when atmosphere of multiculturalism is perceived as high, ethnic belonging is positively related to job satisfaction, but is negatively related to complaints to supervisors. On the other hand, when the atmosphere is perceived as low, ethnic belonging is positively related to the number of grievances filed. In addition to findings concerning multiculturalism, different kinds of individualism-collectivism values were found to influence grievance behaviours in different ways. Coworker collectivism is positively related to complaints to supervisors, complaints to coworkers of same ethnic backgrounds, and the number of grievances filed. On the other hand, general collectivism is negatively associated with complaints to shop stewards, the number of grievances filed, but is positively associated with feelings of animosity and worry of reprisal after grievances were filed. These findings were discussed in light of research implications for individualism-collectivism, Canadian multiculturalism policy, and diversity management. Furthermore, it is recommended that alternative dispute resolution procedures, which allow disputants to structure the process with freedom, be implemented in multicultural workplaces. iii Table of Contents Abstract ii Table of Contents iv List of Tables vi List of Figures vii Acknowledgement viii Chapter 1 Introduction 1 Chapter 2 Literature Review 4 Models of Grievance Procedure 4 Multiculturalism and Related Studies 19 Chapter 3 Social Psychological Analysis of Multiculturalism in the Workplace 34 Use of the Multicultural Framework in Business Organizations 37 A Social Psychological Framework of Multiculturalism for Studying Grievance Procedure 38 Grievance Behaviours 41 English Ability 46 Ethnic Belonging 48 Atmosphere of Multiculturalism 49 A Universal Human Value: Individualism - Collectivism (I-C) 54 Ethnic Composition 62 Evaluations of the Grievance Procedure 63 Demographic Variables 66 iv Chapter 4 Method 67 Selection of Research Site 67 Procedures 70 Questionnaire for Hypothesis Testing 74 Questionnaire for Convergent and Divergent Validity 82 Chapter 5 Results 87 Validation of Constructs 87 Ethnic Categories 95 Analysis on the Three Components of the Multicultural Framework 101 Analysis on Grievance Behaviours 108 Analysis on the Evaluations of the Grievance Procedure 113 Chapter 6 Discussion and Conclusion 121 Summary of Findings 121 Implications of Findings 130 Limitations 142 Conclusion 145 References 146 Appendix A Interview Guideline 162 Appendix B Cover Letter of the Survey Package 163 Appendix C Questionnaire for Hypothesis-Testing 165 Appendix D Questionnaire for Construct Validation 169 Appendix E Self-identified Ethnic Group, First Language, and Grouping 173 v List of Tables Table 1 Hypotheses of the Study 42 & 88 Table 2 Measurement of Variables (Hypothesis-Testing) 75 Table 3 Measurement of Variables (Construct Validation) 83 Table 4 Characteristics of Different Ethnic Groups 98 Table 5 Grievance Behaviours as a Function of Ethnic Grouping 100 Table 6 Correlations between Variables 103 Table 7 Regression Analysis on the Three Basic Multiculturalism Components 105 Table 8 Regression and Tobit Analysis on Grievance Behaviours 109 Table 9 Regression Analysis on Evaluations of Grievance Behaviours 115 vi List of Figures Figure 1 A Six-Stage Model of a Typical Grievance Process 16 Figure 2 Four Components of the 1971 Multiculturalism Policy 35 Figure 3 Social Psychological Framework of Multiculturalism for Studying Grievance Procedure 39 vii Acknowledgement I would like to thank my research supervisor, Dr. Brian Bemmels, and two other committee members, Drs. Larry Moore and Ilan Vertinsky, for their advice and directions concerning this thesis. Many people helped me in the data collection process. They include Paul Straszak, Brian Charlton, Robert Jasmanidse, Arlene Ells, and the interviewees and respondents of the Vancouver plant. Lastly, I would like to acknowledge the Centre for Labour and Management Studies in the Faculty of Commerce for providing financial support for this study. viii Chapter 1: Introduction Grievance procedures have been studied in sociology, industrial relations, and organizational behaviour (e.g. Ewing, 1989; Lewin & Peterson, 1988; Lewin, 1993). However, just as in many management studies (Adler & Bartholomew, 1992; Boyacigiller & Adler, 1991), cross-cultural and ethnic differences are seldom the focus. Given the upsurge of ethnic diversity in the workforce (Alderfer & Thomas, 1988: 16; Burke, 1991; Cox, 1993; Jackson et al., 1991; Jamieson & O'Mara, 1991; Trickett, Watts, & Birman, 1994), a study on the grievance process in a multicultural workplace is needed. Contribution of the Study This project contributes to management research both in theory and in practice. Existing models of grievance procedure are theoretically deficient for a multicultural work environment because they fail to incorporate the dynamics of ethnic relations. Their treatment of ethnic differences is a reflection of the melting pot ideology - to explain why minorities fail to become or conform to the mainstream. As in other management studies, the dominant views are assumed to be universal and minority views are exceptions to be explained (Nkomo, 1992: 490; Watts, 1994: 50). When a "general" theory is tested, ethnic differences are reduced to a demographic variable or are simply ignored. As a consequence, any unique minority characteristics are not incorporated into the "general" theory, and the ethnic dynamism characterized by multicultural societies (Berry & Laponce, 1994) is largely ignored. Although the literature on affirmative action programs (e.g., Taylor, 1991), mentoring (e.g., Thomas, 1993), and organizational demographics (e.g., Tsui & O'Reilly, 1989) addresses some issues related to ethnic dynamics, it has limited 1 relevance for grievance behaviours. Comparative labour relations studies (e.g., Haiven, 1994) offer insights into national differences, but they do not address the dynamics of interethnic contacts in the workplace. Accordingly, a study based on a framework of multiculturalism (Berry, 1984) may better our understanding of grievance behaviours in an ethnically diverse workplace, especially when a recent review found few studies of this kind for the workplace (Berry & Laponce, 1994: 13). In practice, academics and consultants have urged managers to make full use of an ethnically diverse workforce (Fernandez, 1993; Foster et al., 1988; Jackson et al., 1991; Jamieson & O'Mara, 1991; Kirchmeyer & McLellan, 1991; Loden & Rosener, 1991; Moore, 1995) . However, the role of grievance procedures for this purpose has been rarely explored. Although several journals have devoted special issues to ethnic differences in conflict resolution (e.g., LeResche, 1992; Donnellon & Kolb, 1994; Duryea, 1993), the papers have only indirect relevance to grievance procedures. This study may provide useful information for management in this aspect. This study may also provide useful information for unions. Changes in the ethnic composition of the workforce mean that traditional sources of members will decrease, while new memberships will increasingly come from women, ethnic minorities, and disabled workers (Defreitas, 1993; Ray, 1994). To survive, unions need to recruit these traditionally non-organized groups by serving their interests. In addition, unionism as a social movement is obliged to involve these groups into the labour movement (Noble, 1993). Together, it means not only recruiting them as members, but also empowering them in the union and the workplace (Ray, 1994). A step towards this end is to provide an adequate system for them to 2 deal with their dissatisfaction and exercise their rights in the workplace. This study may also offer information on managing a grievance system for this purpose. Overview of the Manuscript This project employs a framework of multiculturalism to analyze various grievance behaviours in an ethnically diverse organization. For reasons detailed below, the study focuses on the informal part of the grievance process, beginning from the point where workers experience dissatisfaction, and running through the informal discussion of potential grievances between workers, supervisors, and stewards up to the filing of grievances. The formal settlement of grievances, which begins with grievance filing and ends in arbitration, is not studied. The remainder of this manuscript is organized as follows. Models of grievance procedures and studies relating demographic variables to grievance behaviours are reviewed in Chapter 2. The review is extended to multiculturalism, ethnicity, and organizational diversity management so as to lay down the theoretical basis for hypotheses tested in an empirical study. Chapter 3 introduces a social psychological framework multiculturalism and delineates hypotheses derived from this framework. Chapter 4 presents the method of the empirical study. Chapter 5 reports the results of statistical analysis on the data. Finally, Chapter 6 summarizes the findings and discusses their implications for research and application. 3 Chapter 2: Literature Review This review is divided into two sections. Models of grievance procedure and demographic influences on grievance behaviours will be reviewed first. Then, in the second section, literature concerning multiculturalism and its related topics, including ethnicity and organizational diversity management, will be discussed. These efforts are an attempt to define the scope of this project and to position it within the existing body of knowledge. Models of Grievance Procedure One recurrent criticism stated by several authors on grievance procedures is that this area of research has been atheoretical and data-driven. Fortunately, since the proclaiming of this theory crisis (Dalton & Todor, 1982; Gordon & Miller, 1984), a number of papers have filled the vacuum with theories and models that address the grievance process from different perspectives (for a review, see Peterson, 1992). Instead of duplicating Peterson's efforts, I chose to cover studies that formed the basis of my understanding of this topic, along with a few publications which Peterson (1992) did not cover.1 After the discussion of these models, the literature pertinent to demographic influences on the grievance process will be discussed. A General Framework of Conflict Intervention Procedures Sheppard (1984) discusses a general framework which describes and explains behaviours of a third-party when intervening in a dispute. It is assumed that the !Most conflict resolution models are not covered here due to the excellent review by Lewin (1993). The comprehensive coverage by James (1992) also makes reviewing cross-cultural research on conflict resolution unnecessary. 4 characteristics or nature of conflicts determine the criteria people use in evaluating conflict intervention procedures. Once a third party understands the nature of a conflict, he or she decides their criteria for evaluating available procedures and tries to administer a procedure that maximizes the set of criteria. The nature of a conflict is defined by (1) the qualities of disputants, such as personality and experience, (2) the causes of the conflict, such as the needs of disputants, (3) the setting of the conflict, such as the nature of constituency relationship, and (4) the involvement of other parties. In general, the nature of a conflict determines the relative importance of the evaluation criteria, and as intervention procedures differ in their power to optimize different evaluation criteria, their effectiveness in resolving a particular conflict is not the same. The evaluation criteria are related to either the procedure or the outcome of it. These procedural and outcome criteria are concerned with one of the four substantive issues -fairness, participant satisfaction, effectiveness, and efficiency. Hence, there are criteria for procedural fairness, such as the level of disputant control and the protection of individual rights; there are also criteria for outcome satisfaction, such as the benefit of outcome for participants and the level of disputant animosity. Sheppard (1984) listed twenty-one criteria. After figuring out the characteristics of the conflict and the resulting criteria of effectiveness, the third party will choose an adequate intervention procedure. Intervention procedures differ in the forms of control a third-party has and his or her timing in using the control. Four forms of control are possible: process control (how disputants interact during dispute resolution), content control (the substantive resolution of the dispute itself), motivation 5 control (the source of power the third party uses to influence the disputants, e.g., persuasion, legitimate authority, threats and promises), and control used only at the request of any one of the disputants. Any one or more of these forms of control can be used at different times during the intervention. The timing can be conceived of as four stages of an intervention, namely definition, discussion, alternative selection, and reconciliation. Crossing the four criteria with the four stages, we have sixteen cells. In theory, any two procedures may differ from one another in any one of these sixteen cells. Stated differently, behaviours of third parties are different in the form of control exerted in the four stages of an intervention procedure, and these differences differentiate one procedure from others. Two examples will clarify this notion. In an adversary arbitration the third party has request and motivational control in all the four stages, but he or she has only process and content control in the alternative selection and reconciliation stages. In other words, disputants in an adversary adjudication can present evidence at their own will, but have no control on the final decision of the adjudication. On the other hand, the third party in a typical mediation has all four types of control of the first three stages but has no control in the reconciliation stage. Hence, disputants are free to choose the solution in mediation, and the third party is only left with the control on enforcing the final decision. According to this framework, a typical grievance procedure can be understood as a special case of an intervention procedure as the workplace constitutes a particular combination of setting, involvement of third parties, and conflicts. The framework, thus, has two implications for studying grievance procedures. First, the notion that intervention procedures 6 have four stages wherein a third party may possess different forms of control is useful in understanding the formal process of grievance procedures. In addition, the criteria for evaluating an intervention procedure will be useful in studying how minority workers perceive and assess existing grievance procedures in the workplace. A System Model of Grievance Procedures Lewin and Peterson (1988) conducted a large scale study on grievance procedures in unionized organizations of four industrial sectors. Their main objectives were to evaluate the effectiveness of grievance procedures with behavioral and attitudinal measures and to investigate post-grievance resolution behaviours, which used to be a neglected area. They put together a comprehensive model of grievance procedures to guide their conceptual thinking and their empirical investigation. Their model posits that the structure, functions, and utilization of the grievance procedure are determined by environmental, organizational, and individual variables. In turn, the use of the grievance procedure affects the labour relations outcomes, with the major variable being grievance resolution. In other words, the use of grievance procedure is hypothesized to mediate the effects of the environmental, organizational, and individual variables on grievance resolution. Unlike other models, this one does not end with grievance resolution. It goes on to explain how grievance resolution together with environmental characteristics affects post-grievance behaviours, such as absenteeism and performance, and how post-grievance behaviours will eventually be fed back to influence the collective bargaining. Several findings are of interest. First, procedural effectiveness was predicted by 7 sixteen variables, which were grouped into four clusters - management characteristics, union characteristics, labour-management characteristics, and the characteristics of the grievance procedure. Although the predictability of the variables varied, several of them were consistent predictors, including management and union policies to take certain grievances as far as possible through the procedure, a union policy of committing grievances to writing, adversarial labour relations, the supervisors' knowledge of the grievance procedures, and the use of expedited arbitration. Second, industry was a significant moderating variable when studying grievance procedures. This is shown in the fact that the effectiveness of procedures, the importance of grievance issues, and the rate of grievances varied across industries. Third, the frequencies of grievance filing and those of specific grievance issues filed varied with demographic variables, including race, age, sex, education, and work experience. Further, a "chilling" effect was found among people involved in grievances. The performance of the grievant and the supervisors of the grievant was negatively affected by the use of the procedure. Users were more likely to leave the company. Lastly, the grievance process is closely related to collective bargaining. Some issues raised as grievances were dealt with in subsequent collective bargaining. An Employee Decision Making Model of Grievance Procedures Klaas (1989) proposes a model on grievance behaviours which includes both conventional determinants and determinants inherited in the grievance system. In the past these sets of determinants were studied separately. One characteristic of this model is that it focuses on the perceptions, attitudes, and decisions of employees, and ignores those of stewards and supervisors. Therefore, it details the cognitions and behaviours of a potential 8 grievant both before and after filing a grievance, and does not consider the influences of other parties in the process. This model starts by portraying the factors that affect grievance initiation. It is suggested that the characteristics of the workplace and employees, and the perceived instrumentality and inequity of the employees, influence whether one decides to file a grievance. In addition, one's decision also depends on the availability of alternative actions, such as quitting the organization and lowering productivity. If alternative actions are plausible, one may not file a grievance at all. In any case, however, the employee can always file a grievance and engage in alternative actions at the same time. The model takes into consideration the post-grievance behaviours of employees as well. Employees who have filled a grievance are hypothesized to be less likely to commit any disruptive actions, unless they feel as if they have been treated extremely unfairly. If they see fairness concerning the procedure, their feeling of inequity will drop, regardless of the grievance outcomes. Experience of the grievance process will also affect their future experience. Specifically, they will engage in grievance activities less often and alternative activities more often if the grievance experience is negative. In addition, the employees who engage in disruptive behaviours are more likely to face future grievance opportunities than other employees who do not. Al l in all, similar to Lewin and Peterson (1988), this model postulates a grievance process as continuous. Hence, determinants of grievance filing are also determinants of the impact of grievance outcomes; moreover, grievance outcomes are fed back to influence the determinants of grievance filing in the future. This model differs from Lewin and Peterson 9 (1988), however, in suggesting that employees can engage in alternative and disruptive behaviours at any point of the process to either replace or go together with grievance filing. Therefore, while considering the grievance process as continuous, we should study grievance behaviours along with alternative actions of employees. Grievance Activities at Different Stages Chaykowski, Slotsve, and Butler (1992) investigated the determinants of grievance filing and grievance decisions at three different levels of a grievance system. This study is important, as it simultaneously modelled grievance processing and final outcomes at separate levels of the grievance system and thus took into account effects caused by self-selection at each level. In contrast, other studies examined grievance filing and grievance outcomes only at the last level, ignoring selection effects which occurred in the middle levels. The results supported the need to study grievance processing at different levels. First, the results showed that the determinants of grievance filing at level 1 and the determinants of bringing the case to levels 2 and 3 were different. The number of previous grievances filed discouraged.filing at level 1, but not in other levels. Higher pay and group grievances discouraged filing at level 2, and seniority discouraged filing at level 3. Second, the results demonstrated that determinants for grievance outcomes were different across levels. At level 1, the chance of a favourable decision was reduced if a similar grievance had been recently granted, but the reverse was true at level 2. At level 3, no predictors were significant. On the other hand, an analysis which modelled the decision at the final level without using the multi-level analysis showed that other variables, such as the percentage of grievance previously granted and the subject of the grievance, were important, but not those found in 10 the multi-level analysis. In sum, the study suggests that different processes and determinants may operate at different levels of the grievance system. Therefore, it is crucial to understand the grievance process separately at each level of a grievance system. A Comprehensive Model of Grievance Initiation Recently a model of grievance initiation has been developed (Bemmels, 1994; Bemmels, Reshef, & Stratton-Devine, 1991). This model differs from previous ones as it uses the behavioral pattern of workers, stewards, and supervisors as predictors for grievance initiation, rather than using the characteristics of these people and those of the workplace. This model thus provides a detailed postulation of the grievance process. Two empirical tests on the model revealed three intriguing findings. First, various grievance-related behaviours of supervisor, employee and steward, such as consideration and work emphasis of supervisors, complaint behaviour of employees, and informal conflict resolution of stewards, determined the grievance rate of work groups. The interrelated nature of the grievance-related behaviours of the parties were shown in two aspects. First, the grievance-related behaviour of any one of the three parties could be affected by those of the others. Also, in accounting for grievance rate, the variables concerning employees explained away the effects of those concerning supervisors. This implied that the behaviours of the two parties were interrelated and jointly affected grievance initiation. In the workplace, interactions between the three parties proved to be important, as informal settlement by stewards significantly reduced the grievance rate of work groups. Second, Bemmels (1994) found that grievance rate varied across industries, although 11 Bemmels et al. (1991) did not find these results, probably due to a smaller sample. Nonetheless, as shown in other studies (Lewin & Peterson, 1988), industrial variations in grievance rate were attributable to industrial differences not captured by the model. Lastly, the initiation process differed across grievance issues. Specifically, parties differed in the kinds of grievances they filed. For example, stewards raised more grievances on job descriptions, whereas employees raised more grievances on discipline, and work groups composed of women raised more grievances concerning discrimination and work rules. Several implications can be drawn from the findings of Bemmels and his associates' studies. First, their works underscore the intertwined behavioral patterns of employees, supervisors and stewards in initiating grievances. Especially, they found that informal settlement by stewards reduced the number of grievances filed. Their work proved the importance of considering the interaction of workers, stewards, and supervisors in the grievance process. This idea contrasts with other models (e.g., Klaas, 1989; Karambayya & Brett, 1989) which are concerned with only one party in the grievance process. Moreover, their works found that women as a minority group were more likely to file discrimination grievances. This finding is in line with findings that ethnic workers as another kind of minority also raise more discrimination grievances (Lewin & Peterson, 1988). Lastly, industrial variations again proved to be important in accounting for grievance behaviours. Demographic Attributes and Grievance Behaviours Demographic variables of individuals have long been studied in organizational studies (e.g. Pfeffer, 1983). Research on grievance procedures has paid attention to them as well (e.g. Bemmels, 1988), and a number of researchers have surveyed demographic influences on 12 grievance behaviours (Allen & Keaveny, 1985: 521; Labig & Greer, 1988; Lewin & Peterson, 1988). Findings are generally inconsistent, although some variables yielded relatively consistent results. These consistent and inconsistent findings are summarized below. Grievants are also found to be younger and have higher skill positions. Allen and Keaveny (1985) reviewed studies in late 60s and 70s and found that age had either no effect or negative correlation with grievance filing. They also found that wage, usually associated with better skills, had either no effects or positive correlation with grievance filing. Occupations may also make a difference. It was found that engineers and faculty members are reluctant to file grievances, as opposed to blue-collar workers. Gender and marital status generally cause no differences, although Lewin and Peterson (1988) and Bemmels (1994) found that men were more likely than women to file grievances. As opposed to the above, findings concerning education, job tenure, income, and union participation are inconsistent, and findings concerning race vary across studies as well. Whites filed grievances more frequently than minorities during the late 1940's and early 50's, but studies conducted since the beginning of the Civil Rights movement have found either no difference between races or a tendency for minorities to file grievances (Labig & Greer, 1988: 15). The filing rate of various ethnic minorities differ among themselves as well. Lewin and Peterson (1988) reported that blacks filed more grievances than whites but Asians, Hispanics and others filed less. Unfortunately, despite these inconsistent findings, race differences per se were usually not treated as theoretically important (Nkomo, 1991), and seldom did researchers use a theory to explain the contradictions. 13 Implications of the Review As a whole, the review identifies several issues pertaining to the grievance process. First, the grievance process progresses in stages, each involving different players, events and activities. The decisions and events in one stage influence those in the next and so on. This is true not only in the formal processes between high-level union and management officials but also in the informal processes among workers, stewards, and supervisors. Therefore, a model of grievance procedures may postulate the grievance process analytically as several discrete stages. Indeed, a multi-stage model is essential, for various social processes occur at different stages of the grievance process (Bemmels et al., 1991; Chaykowski, Slotsve, & Butler, 1992). Second, the grievance process should be considered as an open, continuous process. Previous resolutions affect the evaluation and subsequent usage of the grievance procedures (Klaas, 1989; Lewin & Peterson, 1988). People are less willing to use the procedure if grievants are subject to reprisal or if decisions are not fair. In addition, filing a grievance is only one of the workers' possible reactions to dissatisfactions (Klaas, 1989). People may commit sabotage, and they may complain about their dissatisfaction to friends rather than to stewards and supervisors. The implications are that workers' evaluations of the grievance procedure and their alternative actions to grievance filing should be taken into account in a model concerning the grievance process. Third, the review on demographics and grievance behaviours reveals intriguing but inconsistent ethnic differences. These results, however, come from studies that treat ethnicity as a demographic variable without a theoretical basis for ethnicity. Unfortunately, 14 inconsistent results were usually regarded as trivial by mainstream researchers who dismissed ethnic differences. To ameliorate the situation, theories are needed to make sense of these findings and advance our understanding of ethnicity in the grievance process. Lastly, when researchers draw conclusions about grievance procedures, they should take note of the industry from which they draw their samples. Several studies have found industry to be an important moderating variable to grievance behaviours. This is probably because industry serves as a proxy for differences in HRM practices, environment, and job nature in many of these studies (Lewin & Peterson, 1988). A Model of Grievance Processing The model used for this study incorporates the implications discussed above. It is not supposed to be comprehensive, but serves to illustrate a typical grievance process. As shown in Figure 1, this model is composed of six stages. The first stage is the grievance discovering stage. Grievance discovering is different from grievance filing as it only concerns whether a worker feels dissatisfaction or whether the worker feels being mistreated by the management. Many of these dissatisfactions may disappear as time elapses. Moreover, if the issues are not serious, workers may not take any action at all. If workers are experiencing dissatisfaction, workers may consider voicing it to other people and step into the grievance raising stage. Depending on the circumstances, they may talk about their unhappiness to their family, their coworkers, union stewards, or their supervisors. They may express their dissatisfaction for many reasons, such as seeking social comparison and social support, asking stewards for advice, or complaining to their 15 Figure 1: A Six-Stage Model of Typical Grievance Process Stage Grievance Discovery Grievance Raising Informal Settlement Grievance Processing Grievance Reconciliation Post-Grievance ON Concerns at Various Stages definition of conflict and grievance different working conditions targets of expressing dissatisfaction, such as peers, stewards, supervisors purpose of expressing dissatisfaction, e.g. social support, complaint, changing the situations Informal & formal procedures , relationships between workers, steward & supervisor , flexibility & suitability of the grievance procedure as evaluated by workers . process control grievants like to have In presenting arguments and evaluatng alternatives . reconciling parties with solution . enforcing decisions . hearing appeals , aftermath: exit voice, loyalty, & neglect , reprisal supervisors in the hope of changing the conditions. Simply put, this grievance raising stage starts when workers decide not to keep their dissatisfaction private, and voice it as a way of relieving psychological stress or changing the work conditions. Once workers decide to deal with their dissatisfaction, they have proceeded to the informal settlement stage. Resolving grievances usually goes through informal and formal processes. The informal process occurs when a worker expresses his or her dissatisfaction to the supervisor either alone or together with a steward (Nash, 1983: 14). This informal settlement is crucial because the three parties may well resolve the grievance without any formal processing. As written documents seldom govern this informal process, the three parties can interact flexibly with ample freedom. As the parties enter the formal process, they start discussing various aspects of the grievance. This marks the beginning of the formal processing stage. The formal process is usually structured by the union and the management, and proceeds in accordance with documented procedures. The degree of complexity and formality of the procedure vary across organizations. Nevertheless, as the procedures are usually defined in documents, grievants can exercise little freedom in selecting their procedures. For example, various forms of process control, as suggested by Sheppard (1984: 155), may only be exercised by the parties in a limited manner. In any case the parties at this stage are allowed to present relevant information, clarify information and arguments, and give arguments for each alternative. The parties may refer their case to previous cases (Knight, 1986) and use different tactics to achieve a favourable decision (Martin & Peterson, 1987). The grievance reconciliation stage is closely related to the grievance processing stage, 17 although resolving a grievance does not necessarily require this stage. The union and the management may find a solution or drop the case without involving a third party. However, if a third party, such as an arbitrator or a mediator, is involved, he or she will reconcile a grievant to solutions, enforce the decision, and hear appeals from the disputing parties. Obviously, as organizations usually have a multi-level grievance system, it is possible that the parties will go through several stages of grievance processing before they finally enter the grievance reconciliation stage. The post-grievance stage completes the process. Sheppard (1984) did not have this stage in his model, probably because he was more inclined towards one-shot legal cases. Legal cases do not necessarily involve a long-term relationship between the disputants, so the aftermath of resolution is not a major concern. However, as shown in Lewin and Peterson (1988) and argued by Klaas (1988), post-grievance behaviours and consequences affect the use of grievance procedures and even the collective bargaining and management of the organization. The Empirical Study This model analytically divides the processing of a grievance into six stages. The first three stages and the last three should be studied separately, because collective agreement merely shapes the general context and few formal procedures in the contract exist to confine the social behaviours of workers, stewards and supervisors in the first three stages - grievance discovery, grievance raising, and informal settlement. As a result, workers, stewards and supervisors are actively involved in settling grievances in these stages. In contrast, the last three stages - formal processing, grievance reconciliation, and post-grievance - are usually 18 structured in collective agreement and handled by higher-level union or management officials. The participation of workers, stewards and supervisors is passive and minimal. Thus, if interethnic dynamics influences grievance behaviour at all, we should be able to observe its effect in the first three stages. This project studies grievance behaviours in the first three stages and will not examine those in the last three stages. A number of hypotheses were developed to be tested in the empirical study. They will be presented after the theoretical basis of multiculturalism is laid down. Multiculturalism and Related Research Between 1820 and 1920, as many as thirty-three million Europeans migrated to North America. Continual conflicts between immigrants of different ethnic groups shocked the wealthy new world, where people prided themselves as immigrants who welcomed other immigrants. It was worried that ethnic fractions could eventually tear the nation apart, and was amidst this societal upheaval that Robert E. Park proposed the race-relations theory to make sense of what happened in the process of acculturation (Lambert & Taylor, 1990). The theory was stimulated by a play in New York, entitled The Melting Pot, in which America was depicted as a superior state society due to a "fusion of all the races (Lambert & Taylor, 1990: 26)." This theory suggests that the interactions between the majority and minority start with initial contact, and continue through competition for resources and conflicts, to the gradual accommodation of the minority to existing social institutions, and eventually assimilation (Ujimoto, 1990), which is defined as "a process of interpenetration and fusion in which persons and groups acquire the memories, sentiments and attitudes of 19 other persons or groups, and by sharing their experience and history, are incorporated with them in a common cultural life (cited in Lambert & Taylor, 1990: 25)." Despite the wish of Park and later generations of Americans, assimilation has never or only partially occurred in American society (Reitz & Breton, 1994; Borjas, 1994). One piece of counter-evidence is the "pluralism"observed by Milton M . Gordon (cited in Lambert & Taylor, 1990: 27). Gordon regarded America as "pluralistic," as he observed that the process of assimilation had to proceed through several stages, such as "cultural" assimilation and "structural" assimilation. Different ethnic groups at any one point in time functioned at different stages, while some did not even seem to move at all. Another criticism comes from Lieberson (1985), who pointed out that assimilation occurred only when the differential relationships between the dominant and subordinated groups were relatively stable, so that the subordinated groups would gradually move to become subgroups of the dominant group. Conflicts might arise when the dominant ideology of racial superiority was challenged by the supposedly inferior racial minority. He suggested that whether or not assimilation or conflict occurred depended on particular migrant groups and the conditions when they entered (for other criticisms, see Ujimoto, 1990; Triandis et al., 1986; Lee & Tse, 1991). Canada had similar conflicts among different ethnic groups, although the divergent history of two countries leads to different concerns. Ethnic conflicts began in both countries when European explorers came into contact with the Indian natives or aboriginal groups. Moreover, because of the colonial ideology of the time, both countries discriminated and subordinated subsequent immigrants from Asia, such as Japanese and Chinese (Ujimoto, 20 1990). However, Canada and the U.S. do have many differences in terms of interethnic contact. The primary issue in the U.S. is the incorporation of blacks descended from a large African slave population. The issue is primarily economic, and the recurring question is how to reduce massive polarization of wealth and social status. On the other hand, the primary issue of Canada, from its conquest by the British to today, has been the accommodation of French-speaking and English-speaking Canadians. It is largely cultural and political rather than economic. The problem is to design a political system to allow French language and culture to thrive within dominated English Canadians and their culture (Reitz & Breton, 1994: 3). In short, the issue in the U.S. is to provide full institutional participation for blacks whereas the issue in Canada is the accommodation of two institutional systems based on French and English cultures (Breton, 1978). Despite her differences from the U.S., Canada had policies that followed the assimilation ideology, until counter-evidence finally changed the tide when a policy of multiculturalism was inaugurated by Prime Minister Pierre E. Trudeau in 1971.2 The policy rejected the notion of a single culture as commensurate with national identity and prosperity and sought to take into consideration the contribution made by other ethnic groups to the cultural enrichment of Canada. Federal assistance was made available for maintaining 2The political context of this policy should not be overlooked. In the first place, the policy was established in the framework of two official languages, French and English, in light of the report of the Royal Commission on Bilingualism and Biculturalism in 1969. It was also an attempt to establish a unique Canadian identity with respect to the U.S., defuse the threat of American-style race riots, shore up political strength in Ontario, and neutralize prairie-province grievances (Elliot & Flears, 1990: 64). 21 the cultural identity of minorities, fostering full and equal involvement in economic and social institutions, and encouraging harmonious links through elimination of discrimination at all levels of society (Elliott & Fleras, 1990). In 1988, the policy was given a legislative basis by the Multicultural Act together with more money being earmarked in support of ethnic activities and research. Since 1971, research on Canadian multiculturalism has proliferated (Berry & Laponce, 1994: 13). The following review on ethnicity, multiculturalism, and organizational diversity management aims to outline areas relevant to this project. Ethnicity Ethnicity and race. Multiculturalism is more about ethnic groups than racial groups. Race was used by physical anthropologists and biologists simply as a way of categorizing groups of people into phenotypes that were statistically identifiable. The characteristics chosen bore no relation to culture, personality, or social behaviours. Overtime, however, race has been used to connote meanings including (1) genotypical traits, such as "Mongoloid race," (2) cultural groups, such as the "French race," (3) the human species, and (4) groups different in innate physical characteristics, such as "white" and "black." Although the last meaning is the most popular, race is usually understood as carrying cultural meanings. Due to these multiple interpretations, the meaning of race is not well-defined. On the other hand, ethnicity refers exclusively to the aspects of culture (Ujimoto, 1990). It refers to "that part of an individual's self-concept which derives from his knowledge of his membership of a social group together with the value and emotional significance attached to that membership (Tajfel & Turner, 1979)." In essence, members of 22 an ethnic group share a sense of identity based on descent, language, religion, tradition, and other common experiences (Aboud, 1981; L i , 1990: 5). Emphasizing the cultural aspects, ethnicity has undergone four major conceptualizations in the past two decades (Isajiw, 1993). Four schools of ethnicity. The primordialist school defines ethnicity as something inherited, deriving from the kin-and-clan-structure of human society and hence something fixed and unchangeable. Its validity is, however, seriously challenged by the other schools. The epiphenomenon school suggests that ethnicity is created and maintained by an uneven economy and economic exploitation, and argues that the economic structure is divided into the centre and periphery sectors. Being exploited and forced to do marginal jobs, immigrants and minorities are rounded up in the periphery sector where they develop their own solidarity and maintain their culture. This argument implies that observed ethnic differences are actually class and socioeconomic differences. The situational school suggests that ethnicity is chosen by individuals for their self-interest in a given situation. Therefore, it refers to the individual's rational ascription of ethnic identity to organize the ethnic meaning of his or her social relationships in reference to the requirements of various social situations. Ethnic identification is hence a calculated decision for gaining advantages in specific situations. The subjective school has emerged to explain cases where adherence to an ethnic group cannot be explained by instrumental reasons alone. It emphasizes that ethnicity is a social-psychological reality or a matter of perception of "us" and "them" rather than something objective or "out there." However, this does not mean that all "subjectivists" reject all objective aspects of ethnicity; they merely tend to emphasize socio-psychological experience. An important approach of the subjective school is social constructionism. 23 Constructionists regard ethnicity as emerging, being negotiated and constructed in everyday living - a process which continues to unfold (e.g., Omi & Winant, 1986). It keeps changing after people migrate and are brought into contact with other populations (Berry, 1980). It also changes as ethnic identity unfolds through a life time (Phinney, 1990: 502). Moreover, political events and social demands affect how people define their ethnicity at a particular time period (Li, 1990; Lieberson, 1993). For example, in censuses between 1901-41, non-Europeans in Canada were asked to report "Negro or Mongolian (Chinese or Japanese)," whereas minorities could report either "Chinese," "Black," or other Asian groups, such as "Indian" and "Filipino," in the 1986 Census. Likewise, the split-up of Yugoslavia and Czechoslovakia along ethnic lines has caused the rise of Serbian, Croatian, Czech and Slovak (among others) as ethnic identities in North America (White, Badets, & Renaud, 1993). Measurement of ethnicity. Any empirical study on ethnicity or related areas, such as multiculturalism, requires a measure of ethnicity. Given the noted situational and developmental influences, any measures of ethnicity are better characterized as a state of ethnicity - that is, a person's identification at a given time (Phinney, 1990: 503). A state of ethnicity is first measured by soliciting an ethnic label - a group one uses for defining oneself as a member of a particular ethnic group. Using a single label is to a certain extent problematic, as people with parents of different ethnic groups may define themselves using the ethnic label of either their father, their mother, or a combination of both (White, Badets, & Renaud, 1993). This problem is particularly acute when respondents can only choose from a list of given labels. Alternatively, a recent research approach to measure ethnicity requires people to label themselves in an open-ended question and to state the ethnic 24 backgrounds of their parents at the same time. This allows for cases with mixed parental backgrounds and for an assessment of the implications of differing self-labels with respect to the context of the studies (Phinney, 1992: 159). People may label themselves as a member of a particular ethnic group and yet may not have a strong sense of belonging to the group (Phinney, 1990; Dreidge, 1989). Therefore, the second step to measure ethnicity is to assess the strength of identification. To do so, it is useful to differentiate ethnic identity into symbolic and behavioral components (Dreidge, 1989; Isajiw, 1993; Phinney, 1990). Behavioral ethnic identity consists of outward expressions, such as language, networks, and participation in ethnic activities and organizations. Symbolic ethnic identity concerns mainly psychological aspects, such as attachment, bonding feelings, and experience of common fate. Symbolic identity also includes the attitude towards the ethnic group, such as pride, satisfaction, and contentment. These components of ethnic identity are connected (e.g., Phinney, 1990), and alone or together measure the strength of ethnic identification with different degrees of comprehensiveness. Multiculturalism in Canada and its Research Canada is not the only country that promotes multiculturalism. Sweden and Australia, for example, have similar policies as well. Nonetheless, Canadian research on multiculturalism has progressed significantly and examined various aspects of this policy and the meaning behind it. Firstly, it is important to note that multiculturalism is not only a government policy, but also a demographic reality, an ideology, and a social movement (Elliot & Fleras, 1990: 25 63). In reality, multiculturalism is a description of the existence of different ethnic, aboriginal, and immigrant groups in Canadian society. Multiculturalism as ideology refers to a normative direction on "what ought to be." It dictates programs and actions for achieving ideal states of freedom, tolerance, and respect for individuals. As policy, multiculturalism means governmental initiatives to put ideological objectives into practice through programs and funding. In terms of a social movement, multiculturalism is the activities of ethnic groups and immigrants in their efforts to change their social status under the realm of multiculturalism. Additionally, multiculturalism can be maintained in at least two dimensions: social-structural and cultural (Roberts & Clifton, 1990). The cultural dimension refers to psychological qualities, such as shared cognitive and evaluative beliefs, whereas social structure is defined as a persisting and bounded pattern of social relationships. These two dimensions interact with each other and result in four cases. Assimilation occurs when ethnic groups fail to maintain both their cultural and social-structural distinctiveness. Symbolic multiculturalism refers to situations where only cultural distinctiveness is maintained, whereas ritualistic multiculturalism occurs where only social-structural distinctiveness is sustained. In addition, in case an ethnic group achieves both kinds of distinctiveness, however, it has reached institutional multiculturalism. Institutional multiculturalism is, however, difficult to sustain because it entails strict alignment of individual beliefs with social norms, and without separate social and political institutions and enough members, it is impossible for most ethnic groups. In fact, symbolic and ritualistic multiculturalism focusing on ethnic identity, ethnic activities, and other 26 psychological qualities are the real domains of Canadian multicultural policy (Gans, 1979; Berry, 1984; Roberts & Clifton, 1990). Bearing this emphasis in mind, we can proceed to specific research topics.3 Psychological acculturation. This refers to the changes which individuals experience as a result of contacting other cultures. As discussed, traditional thought on acculturation is dominated by the concept of assimilation, which has been challenged by recent views. One of the counter views is parallel to the concept of multiculturalism as discussed above. This view is based on the finding that establishing close interaction with the dominant society does not necessarily require immigrants to give up their ethnic identity. In fact, affiliating with the dominant society and one's ethnic community do not necessarily interfere with each other (Berry et al., 1991: 278). This view implies, therefore, that marginalization and separation can occur when individuals unwillingly lose their ethnic identity or cut themselves off from both the mainstream and ethnic cultures. On the positive side, however, it indicates that integration can happen when individuals retain their ethnic distinctiveness and at the same time participate in the dominant society. Integration is the acculturation mode that benefits both the immigrants and the society, as research has shown that immigrants who have an integration attitude towards acculturation experience less stress in adapting to the new society, and are more capable of contributing to the new society (Berry et al., 1991: 284). Attitude Towards multiculturalism. Research on this topic is extensive, although most has been criticized as lacking consistent question formats and theory underpinning. Despite 3Areas not discussed here but included in Berry and Laponce (1994) are public media, views of Quebecois, education, human rights, and literature. 27 these shortcomings, Kalin and Berry (1994: 315) found that Canadians moderately supported the principle of multiculturalism; on one hand, they did not approve policy that encouraged immigrants to give up their heritage customs, but on the other they expected established ethnic groups and immigrants to accommodate to Canadian culture. In addition, Kalin and Berry found that education, being British (compared to being French), living in urban areas, and having psychological security enhanced the approval of multiculturalism. Attitude towards multiculturalism is, however, not constant over time. Probably because of economic stagnation and an influx of coloured immigrants, the attitude has shifted from maintaining ethnic heritage to the integrity of nation. Spicer (1991) and Paquet (1994) contend that an emphasis on ethnic differences in multiculturalism, even as a quality which characterizes Canada as a nation, reduces the attainment of becoming "Canadian" and hampers the solidarity of the nation. In response, some researchers argue that most Canadians embrace the principles of multiculturalism, but merely disagree with the policy and implementation (e.g. Kalin & Berry, 1994). Indeed, Berry (1991) observed that multicultural programs had moved away from celebrating cultural distinctiveness towards the promotion of an emerging Canadian identity. Furthermore, some programs were cast with a practical quality claiming, for example, that ethnic diversity improved Canada's advantage in trade and soliciting foreign investment (Multiculturalism and Citizenship Canada, 1993). Wage and social mobility. The central question in studies about labour is related to the vertical mosaic (Porter, 1965). Porter argued that the Canadian mosaic was defined chiefly by stratified ethnic groups that differed in social status and economic inequality. A multicultural policy, he argued, maintained walls that stratified minorities and suppressed 28 their upward mobility. Recent studies, however, show that the division of labour along ethnic lines has dramatically reduced (Darroch, 1985; Lautard & Guppy, 1990), although certain stratifications in terms of wage, the penetration to the elite group, and destined jobs still persist (Reitz & Breton, 1994: Chapter 5). The continuous inflow of immigrants who are professionals and entrepreneurs may eventually collapse the vertical mosaic (Lautard & Guppy, 1990). Discrimination. As overt discrimination is outlawed and sanctioned, covert or polite forms of discrimination are evidenced in the workplace. For example, Reitz and Breton (1994: 83) reported a study showing that whites received three times more job offers than blacks, and that blacks were five times more likely than whites to be rejected for jobs for which whites received interview invitations. Minorities also tend to be left aside from white networks which provide important information and social support (e.g., Alderfer & Thomas, 1988; Fernandez, 1981). Language. In Canada, language is both a means for communication and a crucial symbol of ethnic identity, due in large part to the age-old tension between French and British Canadians (Weinfield, 1994). Immigrants are ridiculed for not speaking "proper" English or for speaking ethnic languages in the workplace. Issues like these usually create tension and distancing among different ethnic groups (Blank & Slipp, 1994). Nonetheless, there exist clear economic advantages for immigrants to have a good command of English or French (Weinfield, 1994: 241). Dispute resolution. Ethnic diversity has aggravated many ethnic-related problems, such as miscommunication, in organizations. Traditional ways of resolving these conflicts are 29 ineffective because of the power differential between minorities and majorities, and the focus of these ways to deal with problems is based on individual meritocracy rather than concerns from ethnic groups (Donnellon & Kolb, 1994). Moreover, with regard to basic issues concerning conflict resolution, there are differences between the views of ethnic groups and those of government-sponsored mediation centres. LeResche (1992) found that the causes of conflicts, the definitions of conflicts, and the ways to resolve conflicts were seen differently by Korean Americans and these centres. Korean Americans tended to view conflicts as damaging relationships, whereas the views of the centres revolved around the differences of self-interests of the disputing parties (see also Donohue, 1994). Similarly, five minority groups in Vancouver reckoned that the mediation centres failed to understand the sources of their disputes and to resolve their disputes in culturally sensitive ways. They recommended that the government adopt a flexible procedure in accordance with the needs and values of different ethnic groups (Duryea and Grundison, 1993). Organizational Diversity Management A book Workforce 2000: Work and Workers for the 21st Century (Johnston & Packer, 1987) has triggered interest in studying non-traditional members of the workforces, such as minorities, women, and the disabled. Since then, the globalization of business has accelerated this trend. Publications on these topics can be roughly categorized into the following. Consulting works. These publications typically begin with some evidence about the change of the composition of the workforce and then follow up with explanations as to why affirmative action programs beginning in the 70's failed to realize the potential of minorities. 30 The conclusion drawn is usually that past efforts focused on passively complying with laws rather than proactively integrating different ethnic groups. They then list the barriers against and discourses for gaining the "diversity advantage," and urge managers to capitalize on the opportunities. Although their arguments usually have their theoretical basis, problem-solving rather than theory-development is emphasized (e.g., Thomas, 1990; Henderson, 1994; Loden & Rosener, 1991; Jamieson & O'Mara, 1991; Thiederman, 1991). Ethnographic studies. Instead of listening to consultants' views on organizational diversity, a few works turned to the organizational participants themselves. For instance, Blank and Slipp (1994) interviewed all the members of groups involved in the diversity context, including immigrants, blacks, and whites. Their work reveals a rich account of the situations and issues that are not recognized by the consultants. Likewise, using ethnographic techniques, Thomas (1993) found that relationships between cross-racial mentors and proteges were a dynamic process which depended on the strategies the parties used in dealing with racially sensitive issues. The process was two-way and involved sequences of subtle exchanges between the mentors and proteges. Laboratory experiments. Controlled experiments have been used to study three strands behind organizational diversity management. One is that different ethnic groups do differ in their behaviours. For instance, Cox, Lobel, and McLeod (1989) found that Asians, Blacks, and Hispanics behaved more cooperatively than whites when working together in groups. The second line of lab studies examined how the ethnic composition of work groups affects workers' behaviours. Espinoza and Garza (1985) found that Hispanics acted more competitively when their representation in work groups was low than when it was high (see 31 also Konrad & Gutek, 1987). The third line of research demonstrates that ethnic diversity translates into competitive advantage. Watson, Kumar, and Michaelsen (1993) found that, compared to homogeneous groups, diverse work groups did less well in the beginning, but subsequently caught up to the former in performance and eventually outperformed them in terms of the range of problem-solving perspectives and the numbers of alternatives generated. Their experiment thus confirmed the creative advantage of diverse work groups. Multiculturalism. This line of research is the closest to the approach taken in this project. In both cases, an ethnically diverse organization is compared with a multicultural society, and findings related to the latter are adapted for studying the former. Cox (1991) specified seven criteria for what he called "multicultural organizations" based on Milton M . Gordon's study of the seven aspects of assimilation. Similarly, Cox and Finley-Nickelson (1991) applied Berry's typology of acculturation process and suggested various conditions that would lead to integration as opposed to assimilation, separation, and marginalization. Further, Kossek and Zonia (1993) and Cox (1993) define the concept of "diversity climate." Just as Berry et al. (1977) argued that a positive multicultural attitude is conducive to the functions of society, they maintain that the diversity climate determines various organizational outcomes. Unfortunately, these works, except Kossek and Zonia (1993), do not come with empirical data. Discussion Three implications are in place following the above discussion. First, it affirms the advantage of adapting theories from studies of Canadian multiculturalism to the studying of organizational diversity. Cox and his colleagues pioneered this approach, but provided little 32 data. Other research on organizational diversity seems to ignore this rich vein of resources. Second, Canadian multiculturalism focuses on the symbolic aspects multiculturalism (e.g., Roberts & Clifton, 199), such as ethnic identity, attitudes and beliefs. Consequently, frameworks adapted from this literature for studying organizational diversity will inevitably be more concerned with symbolic aspects. For example, the measurement of major constructs, such as ethnic identification, would be focused on symbolic aspects. Lastly, as interethnic dynamics is a continuous, unfolding process, a cross-sectional study such as this one will provide only a snapshot of this process. Nevertheless, it is an important steps for understanding the dynamic process, for taking snapshots one after another is one of the ways to understand the whole process of ethnic interaction (Lieberson, 1993: 28). 33 Chapter 3: Social Psychological Analysis of Multiculturalism in the Workplace Berry (1984) explicates the Canadian multicultural policy from a social psychological perspective. He identifies psychological elements in the policy and examines the interrelationships among these elements based on social psychological theories. Although his arguments are chiefly grounded in studies concerning schools, neighbourhood, and families, his framework is applicable to the workplace. A modified version of this framework, which will be used for analyzing grievance behaviours in a multicultural workplace, is presented below. First, let us inspect the framework proposed by Berry (1984; see also Berry, 1991). As shown in Figure 2, the framework is composed of four socio-psychological components and their interrelationships. It is constructed with reference to the key sections of the statement to the House by the Prime Minister (1971): A policy of multiculturalism within a bilingual framework commends itself to the Government as the most suitable means of assuring the cultural freedom of Canadians. Such a policy should help to break discriminatory attitudes and cultural jealousies. National unity, if it is to mean anything in the deeply personal sense, must be founded on confidence in one's own individual identity: out to this can grow respect for that of others and a willingness to share ideas, attitudes and assumptions. A vigorous policy of multiculturalism will help create this initial confidence. It can form the base of a society which is based on fair play for all. The Government will support and encourage the various cultures and ethnic groups that give structure and vitality to our society. They will be encouraged to share their cultural expression and values with other Canadians and so contribute to a richer life for all. It is clear that the policy encourages ethnic groups to maintain and develop themselves 34 Figure 2: Four Components of the 1971 Multiculturalism Policy Group maintenance and development C--J8: O o Intergroup contact and sharing Multicultural assumption or ethnocentrism Backlash or admiration Linguistic vehicle for sharing Group acceptance and tolerance c o -if 03 Learning of official languages SOURCE Berry (1984 ]. as distinctive groups within Canadian society; so the first component of the framework is group maintenance. The second component is other-group acceptance, for the fundamental purpose of the policy is to promote intergroup harmony and the mutual acceptance of all groups. The policy also suggests that group development by itself is insufficient to lead to group acceptance, and intergroup contact, the third component, is also required. Fourth, learning of official language is encouraged because intergroup sharing requires a common medium in order to be successful. There are seven interrelationships among the components. The Multicultural Assumption postulates that confidence in one's own individual identity will bring a sense of security or self-esteem to individuals. As a result, they are more willing to respect and accept other groups. Backlash is, however, possible when an ethnic group becomes so successful that members of that group become the subject of resentment or admiration. If this happens, group maintenance would be either heightened or weakened; findings of both reactions are found. A strong sense of group development may nurture either a confidence of sharing with other groups, or an ethnocentric isolation from other groups. Either consequence is possible depending on the situation. The Contact Hypothesis suggests that group contact is positively associated with group acceptance, but it is true only under certain conditions: the groups in contact have similar status, they share common goals, and the contact is voluntary. There are three interrelationships involving the learning of official language component. The Linguistic vitality assumption embodies the fact that learning the official language results in the loss of heritage language and, consequently, the decline in group 36 maintenance. However, bilingualism is possible. Learning of official languages may be conducive to or independent of the learning of heritage language. The Linguistic sharing assumption suggests that a common language is essential for group sharing to happen. The Linguistic accommodation assumption hinges on findings that individuals who accommodate others in language use are likely to impress other people. Empirical support has been found for many parts of the framework, including the contact hypothesis (e.g., Berry et al., 1977) and interrelationships involving the official language learning component (e.g., Bourhis, 1983), although the framework has not been empirically tested as a whole. Despite this weakness, its richness and clear structure qualify it as a platform for directing research in other areas. Use of the Multicultural Framework in Business Organizations Three reasons justify the application of this framework in the business setting. The first reason concerns objectives. The objective of the Canadian multicultural policy is to engender intergroup acceptance and integrate all groups in the society, so that everyone can fully participate in the society. Attaining this objective is likely to benefit the whole society as well as individual members. Just like the Canadian society, a business organization seeks to maximize its gain and hence the benefits of its members. To do so, it is essential to integrate all groups in the organization and empower them to fully participate in various systems of the organizations (Jamieson & O'Mara, 1991). The second reason concerns the different ways of viewing multiculturalism. As reviewed previously, multiculturalism in society can be viewed in four ways: as a demographic reality, an ideology, a policy, and a social movement (Elliot & Flears, 1990). 37 Multiculturalism in business organizations can be viewed in these four ways as well. Multicultural composition is a reality in many Canadian business organizations (e.g., Moore, 1995). The ideal that "more inclusion of persons from cultural backgrounds that differ from the dominant group (Cox, 1991: 38)" is lauded by researchers on organizational diversity. The beliefs that advantages can be gained through accessing ethnic markets and realizing the potential of people with different backgrounds are widely held by consultants (e.g., Loden & Rosener, 1991). The management of many organizations have initiated and implemented policies and programs to recruit minorities, reduce discrimination, celebrate diversity, and utilize diversity (e.g., Jackson et al., 1991). Valuing diversity has also been become a social movement promoted by unions (Defreitas, 1993) and by scholars to promote multiculturalism in the communities and workplace (Donellon & Kolb, 1994; Duryea & Grundison, 1993). The third reason is related to the components of the framework. Berry's framework is ready for use in analyzing the multicultural workplace, as the four components of the framework are as important in business organizations as in other social settings. Language issues are prevalent and always controversial (Blank & Slipp, 1994). Ethnic identity and intergroup contacts and networks have been studied in organizations (e.g., Ibarra, 1994). Accordingly, as Berry's framework fits the business setting in terms of the objectives, its perspectives, and its elements, it is deemed to be useful for the business setting. However, in order to use it in this setting, some modifications and specifications are necessary. A Social Psychological Framework of Multiculturalism for Studying Grievance Procedures Figure 3 shows a framework which is an operationalization of the more general social psychological framework of Berry (1984). It was designed for the business setting and for 38 Figure 3: Social Psychological Framework of Multiculturalism for Studying Grievance Procedure of the studying of grievance procedures, and abstract components were replaced with measurable constructs. Three of the four components - learning of official language, group maintenance, and intergroup contact - have the same role in this framework, as in Berry's. However, learning of official language is replaced with English ability, because it is the most common language in Vancouver where the study took place. Ethnic belonging replaces group maintenance as a specific measurement on ethnic identification. Atmosphere of multiculturalism indicates more closely the need for integrating different groups in the workplace than intergroup contact. Modifications on the other-group acceptance component are less straightforward, as the component is expanded to include four constructs of grievance behaviours, which reflect the functioning of the grievance system. This framework has four additional sets of variables (shaded in the Figure 3) that do not appear in Berry's model. Measures of individualism-collectivism account for individual differences (Hofstede, 1991; Triandis, 1988); ethnic status taps an important societal influence on the components of the models (Lambert & Taylor, 1990); and demographic composition variables assess the influence of the ethnic composition of the organization (Pfeffer, 1983). These three sets of variables serve to enrich and extend Berry's framework. Lastly, the evaluations of the grievance procedure are also included in the framework. This research area is underdeveloped in labour relations (Boroff, 1991; Clark & Gallagher, 1988) and deserves more exploration. Of these four additions, individualism-collectivism and ethnic status were examined with more analytic vigour, as the rich literature allowed me to make specific hypotheses. Research on the evaluations of grievance procedures, however, is not as rich, and is not 40 directly relevant to grievance behaviours or the multiculturalism components, so making specific hypotheses is almost impossible. Their inclusion is chiefly for exploration. Similarly, the largely inconsistent findings concerning demographic variables also make specifying hypotheses difficult for these variables. Although exploring the effects of demographic variables was conducted, this analysis gave way to analyses that carried stronger theoretical meanings, such as those involving multicultural components and individualism-collectivism. The interrelationships in Berry's framework are meaningful in the workplace although modifications were necessary for studying grievance procedures. Other modifications were also made when I tried to incorporate recent research into Berry's framework. In the following I will discuss these interrelationships and specify hypotheses to be tested in the study. This discussion will start from the grievance behaviours, proceeds to the components of the framework, and moves on to the four sets of new variables. Table 1 summarizes these hypotheses. Grievance Behaviours To recall, grievance procedures can be divided into the formal and informal parts, and this study focused on only the three stages of the informal part. Job dissatisfaction and mistreatment of workers by management have been shown to lead to more grievance filing (e.g., Allen & Keaveny, 1985). As complaints precede grievance filing (e.g., Bemmels, 1994), I expect a positive relationship between job dissatisfaction and the frequency of complaining to shop stewards and filing grievances. Furthermore, dissatisfaction may also incite workers to express their dissatisfaction directly to their supervisors and ethnic 41 Table 1: Hypotheses of the Study Grievance Behaviours Hla: Job dissatisfaction is positively related to the frequency of complaining to supervisors. Hlb: Job dissatisfaction is positively related to the frequency of complaining to shop stewards. Hlc: Job dissatisfaction is positively related to the frequency of complaining to coworkers of own ethnic background. Hid: Job dissatisfaction is positively related to the number of grievances filed. H2a: Complaining to supervisors is positively related to the number of grievances filed. H2b: Complaining to shop stewards is positively related to the number of grievances filed. H2c: Complaining to coworkers of the own ethnic background is negatively related to the number of grievances filed. H3: Informal settlement by shop stewards is negatively related to the number of grievances filed. Linguistic Vitality Hypothesis H4: English ability is associated with ethnic belonging, negatively for the minorities but positively for the majorities. Linguistic Sharing Hypothesis H5: English ability is positively related to perceived atmosphere of multiculturalism. Linguistic Accommodation Hypothesis H6a: English ability is positively related to job satisfaction. H6b: English ability is positively related to complaints to supervisors. H6c: English ability is positively related to complaints to shop stewards. H6d: English ability is not related to complaints to coworkers of same ethnic background. H6e: English ability is positively related to the number of grievances filed. 42 Table 1 (Continued): Hypotheses for the Study Ethnic Confidence Hypothesis H7: Ethnic belonging is related positively to perceived atmosphere of multiculturalism for the minorities, but negatively to perceived atmosphere of multiculturalism for the majorities. Primary Multiculturalism Hypothesis H8a: Perceived atmosphere of multiculturalism is positively related to job satisfaction. H8b: Perceived atmosphere of multiculturalism is negatively related to complaints to supervisors. H8c: Perceived atmosphere of multiculturalism is negatively related to complaints to shop stewards. H8d: Perceived atmosphere of multiculturalism is negatively related to complaints to coworkers of own ethnic background. H8e: Perceived atmosphere of multiculturalism is negatively related to the number of grievances filed. Secondary Multiculturalism Hypothesis H9a: When perceived atmosphere of multiculturalism is high, ethnic belonging is positively related to job satisfaction. H9b: When perceived atmosphere of multiculturalism is high, ethnic belonging is negatively related to complaints to supervisors. H9c: When perceived atmosphere of multiculturalism is high, ethnic belonging is negatively related to complaints to shop stewards. H9d: When perceived atmosphere of multiculturalism is high, ethnic belonging is negatively related to complaints to coworkers of own ethnic background. H9e: When perceived atmosphere of multiculturalism is high, ethnic belonging is negatively related to the number of grievances filed. HlOa: When perceived atmosphere of multiculturalism is low, ethnic belonging is negatively related to job satisfaction. HlOb: When perceived atmosphere of multiculturalism is low, ethnic belonging is positively related to complaints to supervisors. HlOc: When perceived atmosphere of multiculturalism is low, ethnic belonging is positively related to complaints to shop stewards. HlOd: When perceived atmosphere of multiculturalism is low, ethnic belonging is positively related to complaints to coworkers of own ethnic background. HlOe: When perceived atmosphere of multiculturalism is low, ethnic belonging is positively related to the number of grievances filed. 43 Table 1 (Continued): Hypotheses for the Study Ethnic Status Hypothesis H l l : Ethnic status is positively associated with ethnic belonging. H12: Ethnic status is positively associated with perceived atmosphere of multiculturalism. Individualism-Collectivism Value Hypothesis H13: Collectivism is positively related to ethnic belonging. H14: Collectivism is positively related to perceived atmosphere of multiculturalism. H15: Collectivism is negatively related to job satisfaction. H16a: General collectivism is negatively related to the frequency of complaining to supervisors. H16b: General collectivism is negatively related to the frequency of complaining to stewards. H16c: General collectivism is negatively related to the number of grievances filed. H17a: Coworker collectivism is positively related to the frequency of complaining to supervisors. H17b: Coworker collectivism is positively related to the frequency of complaining to stewards. H17c: Coworker collectivism is positively related to the number of grievances filed. H18: Collectivism is positively related to the frequency of complaining to coworkers of same ethnic background. H19a: General collectivism is negatively related to perceived instrumentality of grievance procedures. H19b: Coworker collectivism is positively related to perceived instrumentality of grievance procedures. H20a: General collectivism is positively related to perceived animosity after grievance filing. H20b: Coworker collectivism is negatively related to perceived animosity after grievance filing. 44 coworkers for social support. Therefore, the following hypotheses are made: Hla: Job dissatisfaction is positively related to the frequency of complaining to supervisors. Hlb: Job dissatisfaction is positively related to the frequency of complaining to shop stewards. Hlc: Job dissatisfaction is positively related to the frequency of complaining to coworkers of own ethnic background. Hid: Job dissatisfaction is positively related to the number of grievances filed. Complaining to supervisors and stewards are generally precursors to filing grievances. However, discussing unhappiness with coworkers of the same ethnic group may ease the frustration of dissatisfied workers and reduce their eagerness to file grievances. Hence, the second set of hypotheses are as follows: H2a: Complaining to supervisors is positively related to the number of grievances filed. H2b: Complaining to shop stewards is positively related to the number of grievances filed. H2c: Complaining to coworkers of the own ethnic background is negatively related to the number of grievances filed. Bemmels et al. (1991) and Bemmels (1994) found that grievance rates were lower among shop stewards who reported more informal discussion with supervisors. This is because informal discussion was able to solve some of the problems in the workplace without going to the formal grievance processes. Hence, it is expected that more informal settlement by stewards may reduce the number of grievances filed: H3: Informal settlement by shop stewards is negatively related to the number of grievances filed. 45 English Ability As depicted in Figure 3, the linguistic ability of employees is expected to affect ethnic belonging, and perceived atmosphere of multiculturalism in the workplace, as well as grievance behaviours. There are three hypotheses related to English ability. Linguistic vitality hypothesis. Learning of official languages has been found to be associated with the declining knowledge of the ethnic language (de Vries, 1977). Although education beginning from infancy is effective in producing competent bilinguals (Cummins, 1994), the findings of de Vries may be more applicable to the adult immigrants of this study. If this is the case, one would expect visible minorities, who are mostly non-native speakers, to experience weakened ethnic identification as their English ability improves (de Vries, 1990). In contrast, Anglo Canadians would feel stronger identification the better their English ability, as the language is a symbol of their ethnic group. Canadians who have ancestry of other European origins, such as German and Scandinavian, tend to assimilate with the English Canadian society (Breton et al., 1990; de Vries, 1990), and their languages are more like English than other Asian languages, such as Chinese or Hindi. Having stronger English ability may not reduce substantially, if not enhance, their ethnic identification. Therefore, contrary to visible minorities, the majority, as a whole, may feel a stronger ethnic identification as their English ability improves. The formal hypothesis is as follows: H4: English ability is associated with ethnic belonging, negatively for the minorities but positively for the majorities. Linguistic sharing hypothesis. Language is the medium for human interaction. As 46 one is less likely to learn others' ethnic languages, good English knowledge facilitates exchanging and sharing of cultural activities between different ethnic groups (Berry, 1984). Consequently, because of more involvement and deeper understanding of other groups, individuals with better English ability are expected to perceive stronger atmosphere of multiculturalism in the workplace. H5: English ability is positively related to perceived atmosphere of multiculturalism. Linguistic accommodation hypothesis. All the management functions of the organization, including supervision, training, and filing grievances, are done in English. Although it is possible to use ethnic languages when other parties know the languages, it is unlikely that all the parties that people deal with in the workplace speak their ethnic language. Hence, good English ability is expected to enhance job satisfaction because individuals can understand and communicate better with coworkers and supervisors (Duryea & Grundison, 1993). However, because of the ease of communication, good knowledge of English may also induce more complaints to supervisors, complaints to shop stewards, and the filing of grievances. In contrast, discussing dissatisfaction with ethnic coworkers is not expected to be affected by English ability because individuals can use their own ethnic language. Accordingly, a set of hypotheses are expected: H6a: English ability is positively related to job satisfaction. H6b: English ability is positively related to complaints to supervisors. H6c: English ability is positively related to complaints to shop stewards. H6d: English ability is not related to complaints to coworkers of own ethnic background. H6e: English ability is positively related to the number of grievances filed. 47 Ethnic Belonging One major goal of promoting multiculturalism is to nurture a sense of confidence among the members of the society and, as a consequence, to create more intergroup sharing and cooperation. However, Berry (1984) suspected that heightened ethnic belonging might produce ethnic complacency, which resulted in intergroup isolation and competition as suggested by the social identity theory (Tajfel & Turner, 1979; Kramer, 1991). Ethnic confidence hypothesis. Here I argue that ethnic confidence and ethnic complacency are both possible depending on whether the individual is a minority or a majority. The majority are regarded as socially superior in intergroup contact by other ethnic groups. Kalin and Berry (1994: 299) documents findings on the general population in Canada that British and French Canadians were evaluated more positively than were visible minorities in terms of, for example, being friends, neighbours, and relatives. In addition, their findings show that Canadians of other European descents were ranked between British and French Canadians and the visible minorities. In line with these findings, Goldstein (1985) found that perceived ethnic prestige was found to be in the above order: French and British Canadian, Canadians of European descents, and visible minorities. For the majority group, then, a combination of high status and strong ethnic belonging may result in ethnic complacency leading to intergroup discrimination and negative interactions in the workplace (Tajfel & Turner, 1979; Kramer, 1991). On the other hand, the minorities have less social status than the majorities. Feeling of belonging to their ethnic group, thus, gives them the security and tenacity to interact and exchange with other ethnic groups. Therefore, the following hypothesis is proposed: 48 H7: Ethnic belonging is related positively to perceived atmosphere of multiculturalism for the minorities, but negatively to perceived atmosphere of multiculturalism for the majorities. Atmosphere of Multiculturalism Whether group maintenance and intergroup contact enhance outgroup acceptance is another major concern in the multicultural framework of Berry (1984). He argued that more intergroup contact under benign conditions (dubbed Contact hypothesis), and that ethnic identification that led to psychological security (dubbed Multicultural assumption) would be useful. This study takes a slightly different position from Berry in light of the literature on organizational diversity management and situational ethnicity. Primary multiculturalism hypothesis. In response to the demographic change of the workforce and globalization of business, many organizations have moved beyond training on reducing discrimination and implementing affirmative action programs (e.g., Fernandez, 1981; Alderfer & Thomas, 1988) to valuing ethnic diversity (e.g., Jamieson & O'Mara, 1991). Various methods and interventions have been suggested to create multicultural organizations (Cox, 1991), to integrate the informal networks of different ethnic groups (Ibarra, 1992), and to promote cooperation between the members of different ethnic groups (Pettigrew & Martin, 1988). All these efforts are believed to produce a multicultural atmosphere that strengthens the competitiveness of the organization (Fernandez, 1993) and benefits the members of all groups (Jackson et al., 1991; Kossek & Zonia, 1993). Multicultural atmosphere could be characterized by two aspects (Berry et al., 1991: 287). First is a social network that provides support to those entering into the organization. This network helps remove situations in which minorities are cut off from the mainstream 49 network (Lincoln & Miller, 1979; Morrison & von Glinow, 1990). Second is the greater tolerance for and acceptance of cultural diversity. Having greater tolerance may change white managers' expectation that minorities ought to be assimilated and conform to their ethnocentric ways of doing things (Cox, 1991). Accordingly, if a positive atmosphere of multiculturalism is experienced, it is likely that better communication and understanding among organizational members would be achieved. Consequently, organizational members would have higher job satisfaction. Furthermore, better communication and understanding may also reduce complaints to supervisors, stewards and coworkers, and the number of grievances filed. H8a: Perceived atmosphere of multiculturalism is positively related to job satisfaction. H8b: Perceived atmosphere of multiculturalism is negatively related to complaints to supervisors. H8c: Perceived atmosphere of multiculturalism is negatively related to complaints to shop stewards. H8d: Perceived atmosphere of multiculturalism is negatively related to complaints to coworkers of own ethnic background. H8e: Perceived atmosphere of multiculturalism is negatively related to the number of grievances filed. Secondary multiculturalism hypothesis. Whether or not cultural diversity is recognized in the workplace, as mirrored in perceived atmosphere of multiculturalism, may also modify the relationships between ethnic belonging and grievance behaviours. To recall, Berry and his colleagues (e.g., Kalin & Berry, 1994) contended that strong ethnic belonging might 50 breed ethnocentrism and ethnic complacency, which led to ethnic discrimination and conflicts. In addition, they argued that a positive intergroup relationship was possible when ethnic belonging resulted in a sense of security and self-esteem. In face of these opposing possibilities, I would argue that the multicultural atmosphere may determine which one of the two possibilities shows up. Social situations are powerful moderators of human behaviour. For example, in classic studies on social facilitation, researchers showed that a cheering crowd facilitated the performance of workers, but the performance actually dropped when the cheering started to distract the workers (Zajonc, 1965). Research on situational ethnicity shows that social context may determine which one of a person's communal identities or loyalties is useful in a particular situation. This means that people vary their behaviours based on their perceptions of the situation (Okamura, 1981; Trimble, 1989). For example, Hispanic American students cooperated with coworkers when they were the majority in a small group but behaved competitively when they were the minority of the group (Espinoza & Garza, 1985). Stayman and Deshpande (1989) found that Chinese and Mexicans felt stronger ethnic identification when dealing with racially biased clients, and ordered ethnic food instead of mainstream food in a dining gathering. In the workplace a multicultural atmosphere provides a context that allows employees to develop a healthy identity and mutually positive attitudes. As defined above, an atmosphere of multiculturalism features a social network that provides help and support, and an attitude that values differences. Accordingly, in an atmosphere where ethnic groups are viewed as equal and are encouraged to express their uniqueness, a strong and secure identity 51 promotes people's contact with other ethnic groups. Because of the security, they would also communicate well with managers and fit in well with the management system. Consequently, organizational members would have higher job satisfaction, complain less to their supervisors, stewards, and coworkers of own ethnic background, and file fewer grievances. Formally, the following hypotheses are proposed: H9a: When perceived atmosphere of multiculturalism is high, ethnic belonging is positively related to job satisfaction. H9b: When perceived atmosphere of multiculturalism is high, ethnic belonging is negatively related to complaints to supervisors. H9c: When perceived atmosphere of multiculturalism is high, ethnic belonging is negatively related to complaints to shop stewards. H9d: When perceived atmosphere of multiculturalism is high, ethnic belonging is negatively related to complaints to coworkers of own ethnic background. H9e: When perceived atmosphere of multiculturalism is high, ethnic belonging is negatively related to the number of grievances filed. Conversely, if the atmosphere does not value multiculturalism, the reverse would be the case. This means that people withdraw from contacting other groups and develop negative and deprecatory attitudes toward other groups. In this case, stronger ethnic belonging may only reinforce the clinging together to ones' own groups and the development of insecurity. As a consequence, communications and understanding would be thwarted, and organization functions suffer. The following hypotheses are, therefore, proposed based on this discussion: 52 HlOa: When perceived atmosphere of multiculturalism is low, ethnic belonging is negatively related to job satisfaction. HlOb: When perceived atmosphere of multiculturalism is low, ethnic belonging is positively related to complaints to supervisors. HlOc: When perceived atmosphere of multiculturalism is low, ethnic belonging is positively related to complaints to shop stewards. HlOd: When perceived atmosphere of multiculturalism is low, ethnic belonging is positively related to complaints to coworkers of own ethnic background. HlOe: When perceived atmosphere of multiculturalism is low, ethnic belonging is positively related to the number of grievances filed. Ethnic Status Hypothesis Social identity theory (Tajfel & Turner, 1979) states that people strive for a positive group identity by competing with other groups. Brewer (1979) and Messick and Mackie (1989) documented findings that winning an intergroup competition and gaining status usually resulted in a heightened identification with one's group.1 A classic study found that more students wore the T-shirt of their college, a symbol of identification, to affiliate themselves to their school after the football team won a game in an inter-college tournament. The numbers of student wearing the T-shirt were positively correlated with the wining margins (Cialdini et al., 1976). Indeed, most people strive to associate with winners (Cialdini, 1988: 190). Based on these findings, it is possible that people would identify with their ethnic group the 'Ethnic status in this study is defined as whether one's ethnic group is respected by other groups in the society and whether one's ethnic group is discriminated against in the society. 53 higher the status of their ethnic group. This argument leads to the following hypothesis: H l l : Ethnic status is positively associated with ethnic belonging. Ethnic status may also affect one's multicultural attitude. Even in a multicultural workplace, status of ethnic groups are different, and having a generally high status reduces the chance of running into discrimination and racially stressful situations (Berry, 1980). In addition, having less stress and not feeling being discriminated may facilitate people to help others and promote cultural diversity. Indeed Berry et al. (1977) and Lambert and Taylor (1990) found that cultural security of the ethnic group was associated with endorsement of multicultural policies and programs. Intergroup studies also showed that a high and stable social status reduced discrimination and denigration of other groups (e.g., Caddick, 1982). The foregoing discussion leads to this hypothesis: H12: Ethnic status is positively associated with perceived atmosphere of multiculturalism. A Universal Human Value: Individualism - Collectivism (l-O Human value has long been an important topic in organizational studies (for a review, see Pinder & Stackman, 1995). Kluckhohn (1951) defined values as "a conception, explicit or implicit...of the desirable which influences the selection from available modes, means, and ends of action." Rokeach (1968) enjoys more attention in organizational studies, he defines value as "...an enduring belief that a specific mode of conduct or end-state of existence is personally and socially preferable to alternative modes of conduct or end-states." Values have been found to vary across nations (e.g., England, 1975; Connor, Becker, Kakuyama, & Moore, 1993; Peterson & Fuller, 1992; Schwartz, 1992). Hofstede's (1980) study in fifty countries shows that cross-cultural differences exist on four universal value 54 dimensions, which corresponds to basic problems existing in every society. Power distance is concerned with relations to authority; individualism - collectivism is pertained to the relationship between self and the group; masculinity - femininity is related to individuals' concept of being born as a boy or a girl; and uncertainty avoidance is pertinent to ways of dealing with uncertainty. One of the four value dimensions, individualism - collectivism (I-C), will be emphasized in this study to shed light on how individuals may differ within the multicultural framework; something that Berry (1984) ignores. The I-C value construct is explored here for two reasons. First, it has drawn a lot of research attention. It not only has a long intellectual history (Kluckhohn, 1951; Parsons, 1977; Triandis, 1988, 1989), but also induces ample empirical research, for instance, in risk-taking (Watson & Kumar, 1992), job-related attitudes (Bochner & Hesketh, 1994), differences between French and British Canadians (Major et al., 1994), cooperation in work groups (Cox, Lobel, & McLeod, 1991), and others (see Erez & Earley, 1993). In addition, I-C value exists at both the culture and individual level of analysis (Bond, 1988; Leung & Bond, 1989). This means that it is meaningful to discuss and measure individual differences in terms of I-C. In contrast, other value dimensions of Hofstede may only exist at the culture level-of-analysis (Hofstede, Bond, & Luk, 1993). Hofstede (1991: 51) defines "individualism as pertaining to societies in which the ties between individuals are loose: everyone is expected to look after himself or herself and his or her immediate family." On the other hand, "collectivism pertains to societies in which people from birth onwards are integrated into strong, cohesive ingroups, which throughout people's 55 lifetime continue to protect them in exchange for unquestioning loyalty." Triandis (1988) gave several "defining attributes" of the construct. For the sake of explanation, collectivists are contrasted to individualists, although the construct is actually a continuum. Collectivists pay more attention to some identifiable ingroups and behave differently toward members of that group compared to those of the outgroups. Ingroup goals have primacy over individual goals, and individual behaviours are largely regulated by ingroup norms. Hierarchy, harmony and face saving are also important for collectivists, who favour homogeneous ingroups and insist that disagreements should not be known to outgroups. In contrast, individualists perceive a less clear-cut difference between ingroups and outgroups. Individual goals have primacy over ingroup goals, and behaviour is largely regulated by individual preferences and cost-benefit analyses. Moreover, I-C value is a multifaceted construct. Specifically, while a person has a general tendency of I-C, he or she may be more individualistic or collectivistic depending on the targets of concern and the contexts of interaction, such as family, neighbours, coworkers, friends, and others (Hui & Triandis, 1986; Triandis, 1988). In short, one may display collectivistic or individualistic behaviours toward different target persons. Hui (1988: Study 6) showed that friend collectivism correlated positively with willingness of sharing the blame when a friend broke the window of a neighbour, but neighbour collectivism did not correlate with this willingness. However, when one was responsible for breaking the window of the neighbour, neighbour collectivism correlated positively with willingness to acknowledge the blame, but not friend collectivism. In line with this finding, Ear ley (1993) found that individuals who were high in coworker collectivism displayed more social loafing when they 56 worked together with strangers, but more self-sacrifice when they worked with coworkers (see also Leung & Bond, 1984; Mann, 1988). The multifaceted nature of I-C has important consequences for the socio-psychological framework of grievance behaviours. The general I-C value captures the broad individualistic-collectivistic tendency across a large number of target persons outside the workplace, including parents, kin, neighbours, and friends. Its value reflects how people are generally integrated into strong, cohesive ingroups outside the workplace as opposed to as opposed to being independent and loosely tied with others. It also reflects their tendency of putting their personal goals behind the collective goals of these groups (collectivists) as opposed to putting personal goals in front of collective ones (individualists). On the other hand, the coworker I-C value captures individualistic-collectivistic tendency specifically for coworkers in the workplace; its value indicates the extent to which individuals are related to their coworkers and willing to put their goals behind the coworkers' (collectivists) as opposed to being loosely tied with their coworkers and putting personal interests ahead of coworkers'. Given the multifaceted nature of I-C, coworker and general I-C values are expected to exert different influences on other constructs in the framework. I-C value hypothesis. The first hypothesis pertains to ethnic belonging. Self-concept includes personal attributes, such as intelligence and skills, and social attributes, such as knowledge, value, and emotions of membership in a social group (Aboud, 1981; Tajfel, 1978: 63). Markus and Kitayama (1991) document findings that collectivists utilize more group-orientated self-conception than individualists (Americans). For example, Bond and Cheung (1983) found that when asked to finish sentences with the format "I am .", 57 Chinese filled in more social attributes than Americans, who filled in more personal attributes. Accordingly, collectivists regard group membership to be more important than individualists, and so their ethnic belonging is stronger than that of individualists (Gudykunst, 1988). This expectation applies to both I-C values, although general I-C value may exert a stronger influence than coworker I-C. This is because ethnic belonging pertains to ones' belonging across a number of target persons other than merely coworkers in the workplace (Fishbein & Ajzen, 1985). H13: Collectivism is positively related to ethnic belonging. The second hypothesis based on I-C value pertains to atmosphere of multiculturalism. Research shows that collectivists have high tolerance to, and are willing to provide help for their ingroup members (Triandis, 1988, 1989; Markus & Kitayama, 1991). Comparing to individualists, collectivists are more likely to sacrifice for their ingroups and less likely to "shirk" against them (Earley, 1993). They tend to allocate rewards equally among ingroup members (Kim, Park, & Suzuki, 1986), and they also provide more social support to their ingroups (Triandis et al., 1991). Therefore, we would expect collectivism to be related to perceiving a positive atmosphere of multiculturalism. Coworker collectivism may exert a stronger effect because it governs to what extent a person tolerates differences among and provides for their coworkers. On the other hand, general collectivism reflects to what extent a person tolerates differences among and provides for other target persons in general, and its effects on their perceptions in the workplace may be weaker. H14: Collectivism is positively related to perceived atmosphere of multiculturalism. I-C value may affect different aspects of grievance behaviours. For one thing, Erez 58 and Earley (1993) argue that I-C value affects a wide range of organizational behaviour. For instance, motivation techniques for collectivists should focus on collective benefits, whereas those for individualists should emphasize on personal gains. Additionally, collectivists may negotiate with less compromise and recession, for they have a stronger need for positive evaluations from their ingroups. Given that present management practices are established by and for individualistic white employees (e.g., Fernandez, 1993; Jamieson & O'Mara, 1991),2 employees who carry collectivistic values may feel dissatisfied with existing management and have problems in working with them. As Erez and Earley (p. 233) warned, "management practices should be congruent with cultural values in order to be smoothly adopted and to have positive effect on behaviour." Incompatible management practices and expectations may result in more job dissatisfaction for collectivists. It is suspected that coworker I-C may have a stronger effect because of its relevance at the workplace, whereas general I-C is concerned with other target persons in general. H15: Collectivism is negatively related to job satisfaction. Studies on conflict resolution have showed that I-C value may influence grievance behaviours in the workplace. Leung (1987) and Leung and Lind (1986) examined the kinds of conflict resolution procedures preferred by people varying in I-C value. Their findings suggest that collectivists prefer mediation and negotiation to arbitration because collectivists prefer saving face and maintaining harmony to direct confrontation. In contrast, individualists 2Beck and Moore (1985) suggest that Canadian managerial approaches are based on ethnocentric values and attitudes which have evolved out of a North American, largely Caucasian, milieu. See also Moore (1995) for empirical support. Trickett et al. (1994) recited the recognition of diversity in psychology and social sciences in general. 59 prefer arbitration to the other two procedures, for they prefer maintaining control over the process of resolving the conflict (see also Ting-Toomey, 1988). Likewise, comparing the legal culture of Japan and the U.S., Hamilton and Sanders (1992) found that collectivistic Japanese tended to reduce escalation and strive for repairing relationships between disputants, and mediation and negotiation were preferred to litigation. Even when litigation was used, the procedure was inquisitorial rather than adversary. The court listened to reports from neutral parties rather than arguments made by the disputing parties. Therefore, as collectivism is related to avoiding confrontation and striving for harmony, it is expected to correlate negatively with complaints to supervisors and stewards. Complaining to them inevitably results in confrontation and heated argument which people with high collectivism seek to avoid. However, although the aversion of conflict may be positively related to general collectivism, it may be negatively related to coworker collectivism in the workplace. The reason lies in the fact that general and coworker I-C values cast different views on people concerning who are their ingroups and who are not. High general collectivism disposes people to regard parents, friends, and neighbours as their ingroups, and they are willing to sacrifice themselves for these ingroups, but not others. Therefore, facing a potential conflict in the workplace, people high in general collectivism would tend to refrain from the situation because coworkers are not highly regarded as their ingroups, so their predominant response is to preserve harmony. On the other hand, high coworker collectivism means that people strongly regard coworkers as their ingroups, and they would strongly regard supervisors and managers who 60 threaten the interests of their ingroups as their outgroups. As a result, people high in coworker collectivism are willing to sacrifice their own interests for collective interests of coworkers, and would sacrifice their preference for harmony and compete with their outgroups. Bond and Hwang (1986: 262) found that although Chinese (collectivists) generally preferred soothing, relationship-maintaining approaches to reconcile differences, they undertook abrupt and competitive actions against their outgroups. Leung (1988) found that compared to individualists, collectivists were less likely to sue their friends, but more likely to do so to strangers (their outgroups). Moreover, Chan (1991) found that once collectivists found that a friend, their ingroup, double-crossed them in a negotiation, they recategorized their friend as an outgroup and tried hard to take revenge against their friends. The retaliatory reaction of individualists in his study was not as strong. The argument, then, goes that although general collectivism is negatively related to complaints, coworker collectivism is positively related to complaints. It is because the latter value induces workers to get even with supervisors and managers, who are regarded as outgroups when conflicts arise. This arguments leads to two sets of hypotheses pertaining to collectivism and complaint behaviours: H16a: General collectivism is negatively related to the frequency of complaining to supervisors. H16b: General collectivism is negatively related to the frequency of complaining to shop stewards. H16c: General collectivism is negatively related to the number of grievances filed. H17a: Coworker collectivism is positively related to the frequency of complaining to 61 supervisors. H17b: Coworker collectivism is positively related to the frequency of complaining to shop stewards. H17c: Coworker collectivism is positively related to the number of grievances filed. Collectivists cherish the closeness of their ingroups, for their ingroups always give them strong social support. Individualists also receive support from their friends and coworkers, but as evidence stands these groups are not as close as their collectivists' counterparts, and their support is not as strong or valued as highly as that of collectivists (Triandis et al., 1991; Wheeler et al, 1991). Therefore, I expect collectivists to complain to their ethnic coworkers more often than individualists do. Again, the effect of coworker I-C value may be stronger than general I-C value because the former is more relevant to the workplace than the latter. H18: Collectivism is positively related to the frequency of complaining to coworkers of own ethnic background. Ethnic Composition Heterogeneity may be viewed as either positive or negative in societies and organizations. While it is believed to facilitate innovation and enrich life and culture, it may cause problems. Research summarized in Jackson (1992) on organizational demographics (e.g., Pfeffer, 1983) and attraction-selection-attrition model (e.g., Schneider, 1987) shows that diversity in terms of age, tenure, gender, and race increases role ambiguity, absenteeism, and turnover. These negative consequences of diversity are caused by less interpersonal attraction (Tsui & O'Reilly, 1989), less communication (Pfeffer, 1983), heightened social 62 distinction and social boundaries (Konrad & Gutek, 1986), and decline in social status (Tsui, Egan, & O'Reilly, 1992). However, this line of research has its critics, for the effects of demographic composition, like other factors, are limited. For instance, Thomas (1993) found that cross-racial mentoring relationships, though crippled by racial differences, were not necessarily unpleasant and fruitless. If the mentor and mentee approached the sensitive racial problems positively, their relationship could be fruitful compared to pairs who avoided the problems. As the existing literature does not deal with grievance behaviours, it is difficult to propose specific hypothesis concerning how perceived composition of minorities and majorities affects various constructs in the framework. Deriving specific hypothesis is also complicated by the fact that ethnic composition pertains to coworkers, supervisors, and stewards. The composition variables were included merely for exploratory purposes. Evaluations of the Grievance Procedure Knowing how workers evaluate grievance procedures is important for two reasons. The first reason is that union members' evaluations of the grievance procedure is one of the most important factors affecting their union commitment (Clark & Gallagher, 1988). The second reason is that studies on grievance usage have been criticized as not revealing what really makes a grievance procedure effective (Boroff, 1991: 209). Most studies are only about who uses the procedure, who benefits from the procedure, and why users used the procedure. A study on how both users and non-users evaluate the system is also important. There is, unfortunately, a dearth of studies in this area. One exception is Boroff (1991) which studied how criteria, including fairness, reprisal, feedback of grievance results, 63 the favorability of grievance outcome, and adversary confrontation, affected perceived effectiveness of the grievance procedure. She found that all the above criteria were important in predicting effectiveness among workers and managers. Clark and Gallagher (1988) took a different approach. They studied how demographics, work experience, union experience, and competence of stewards3 influenced perceived importance and fairness of the procedure. They found that work experience and competence of stewards exerted positive effects on fairness. Competence of stewards also had some positive effects on the perceived importance of the grievance procedure. This study explores how the key components in the multiculturalism framework affect five aspects of the evaluations of grievance procedure identified in the two studies discussed above: importance, fairness, instrumentality, worry of reprisal, and animosity between supervisors and workers. Feedback of grievance outcomes, though important (Knight, 1986), was not studied, for it generally carried less weight than the above factors in Boroff's model (1991). A few hypotheses follow from previous discussion on how I-C value influences the preference of conflict resolution procedures. A typical grievance procedure is closer to the ideal of individualists rather than that of collectivists. A grievance procedure is usually composed of a few steps, and after initial attempts at informal settlement have failed, the procedures become formal. Written grievances are appealed to higher and higher levels of the union and management until, if necessary, a third-party intervention (Gandz & Whitehead, 3They called this variable attitude towards stewards, but what they measured was actually how competently stewards had provided help in handling grievances for workers. 64 1989).4 This design is confrontational. As disputing parties ask undesirable or embarrassing questions and pitch sharp arguments against each other, escalation and distance between the parties often increase, and damaged relationships are likely to be exacerbated. Therefore, a typical grievance procedure is confrontational. As argued previously, general I-C value and coworker I-C value may exert different effects on how workers evaluate the procedure. It is expected that general collectivism is negatively, and coworker collectivism positively, related to instrumentality. This is because people high in general collectivism tend to see the grievance procedure as causing heated argument, disrupting harmony, and breeding animosity between workers and supervisors. Those high in coworker collectivism would also see the same, but because their main concern is their coworkers, they would regard management as outgroups and compete fiercely to "defeat" the management. Under this circumstance, harmony, though preferred, is replaced by animosity, as people are willing to sacrifice themselves for the interest of their ingroups -coworkers. Based on the above discussion, the following sets of hypotheses are derived concerning instrumentality and animosity: H19a: General collectivism is negatively related to perceived instrumentality of grievance procedures. H19b: Coworker collectivism is positively related to perceived instrumentality of grievance procedures. ^he four companies studied in Lewin and Peterson (1988: 169) had an informal step and two or three more formal steps before third party arbitration. 65 H20a: General collectivism is positively related to perceived animosity after grievance filing. H20b: Coworker collectivism is negatively related to perceived animosity after grievance filing. Demographic Variables Several demographic variables, including gender, age, education, years in Canada, and experience with unions, repeatedly appeared in studies on grievance procedures, ethnic relations, and other organizational studies (see review in Chapter 2). In this study they were tested for their effects on English ability, ethnic belonging, and atmosphere of multiculturalism, because some demographic variables have displayed consistent effects (Berry et al., 1977; Reitz & Breton, 1994). For example, age is negatively, but education is positively related to approval of multiculturalism policy. Education is also positively related to English ability, and years in Canada is negatively related to ethnic identification. In studies on grievance procedure, education and experience with union were found to be negatively related to complaints and grievance filing (e.g., Bemmels, 1994). These demographic variables therefore need to be controlled in the regression models. When they were found to exert insignificant effects on the criterion variables, however, they were dropped from the final models for simplicity and preservation of statistical power. 66 Chapter 4: Method Selection of Research Site The objective of the study is to investigate grievance behaviours using a framework of multiculturalism. The emphasis is on theory-testing rather than population parameters (McGrath, Martin, & Kulka, 1982). Hence, I aimed at securing a site wherein a variety of ethnic groups worked together. This means that random samples gave way to judgmental samples which have a variety of ethnic groups (Lonner & Berry, 1986: 89). To do so I sought an organization with substantial minority staff rather than one whose profiles of workers were representative of British Columbia or Canada. One organization was studied. Confining the study to one organization controls many variables that may interfere with the central thesis of the study. Compared to a study of many unions and organizations, this single-case study controls organizational culture, provincial differences in organizations, demographic differences across organizations, and general economic conditions. The disadvantage is, of course, restriction on generality, as findings of this study may be qualified by the specific circumstances of this study (see Chaykowski, Slotsve, & Butler, 1992). While trading off generality (external validity), I believe that focusing on one organization is appropriate for this study (McGrath, Martin, & Kulka, 1982). Site. The study took place at the Vancouver plant of a Crown corporation in Canada. According to the annual report of 94-95, the corporation employed a total of 62,500 full-time and part-time workers, and had a total revenue of $4.7 billion in 1994-95. Despite the huge revenue, the company suffered a loss of $69 million in 1994-95, but this was a drop from the 67 $270 million loss in 1993-94. The industrial relations of the organization was described as antagonistic, frustrating, and unproductive by researchers who did research in other plants of the corporation in the 80s.1 It was also affirmed by the interviewees of my study. Grievances were backlogged seriously when I was collecting the data in May, 1995. When approached by me inside the plant, some workers rejected the idea of doing a questionnaire and refused to talk to me about the grievance procedure. Some said sarcastically that I had chosen the "right" topic, and they wished me "good luck" in my dissertation. Fortunately, many of the workers were open-minded and were willing to complete the questionnaire. While bad memories linger, many members remarked that the labour relations has improved steadily in the past two years. One supervisor described the present atmosphere as "warm and fuzzy "compared to a few years ago. These changes were credited by some of them to a new director and labour officers. Others thought that the stagnant economy forced the union and management to cooperate. The changes may also be a reflection of a dramatic change in corporate philosophy. The 94-95 annual report says that "this year is an important year ...in response to today's new realities...we make a dramatic shift from an operation-driven corporation to a customer-driven one." It mentioned the importance of employees to the corporation. At one point, it says that "this year is a major milestone in our history, as our corporation reached with the signing of four labour contracts without disruptive or inconvenience to our customers. This comes with a significant reduction [34%] in the number of workplace grievances." xThe books are not cited to maintain the anonymity of the organization. 68 The corporation seems to be committed to retraining their staff and improving the labour relations. It put in $57 million to establish a training facility, and implemented many "progressive" measures in the workplace. One example of these measures is the joint union and management meetings where management consult with the unions about policy changes. Although a shop steward dismissed the meetings as "informing rather than consulting," and "another idea that wastes time," this comes as a surprise for a corporation which was described by both supervisors and union stewards as "autocratic" and "dictatorship-like since the end of WWII." The Vancouver plant has a high proportion of minority and immigrant workers. Although race is not kept in files, stewards and supervisors estimated the proportion of visible minorities, many of them Asians, to be around 25% to 50%. Many of these workers were hired in the 70s when mainstream Canadians were able to leave this conflict-ridden organization and find work elsewhere. Their vacancies were filled by the then newly arrived immigrants. The quality of work of these Asian and minority workers has impressed the management, and they continue to hire them to work for the plant. Despite their significant presence, minority workers are seldom shop stewards or supervisors. A minority steward told me that there was only one other steward of his ethnic background in the 1,200-worker plant. According to him, most minorities do not want to get into trouble. Other minority workers who have the seniority to become supervisors usually apply for transfers, so that they can manage service outlets in the city instead of staying in the plant, which he called a "dump." The staff of the plant work either in-house or outdoor. The in-house staff take care of 69 shipments of commodities, sort them, and prepare them for the outdoor staff for delivery. The plant operates 24-hours a day in three shifts. Every shift spans eight hours with two coffee breaks and a 30-minute meal break. Most of the staff are full-time. Part-timers who usually work in the afternoon and night shifts are called in when needed. The system of calling in "casuals" has constantly been the major source of grievances, over 70% of all grievances. Many full-timers accused management of failing to notify them for over time work. They accused management of doing this not because of negligence, but because of i l l -will and the desire to avoid overtime pay. Other relatively common grievances involve wage and sick leaves. A national union represents the staff of the corporation, and compulsory membership is in place. When the study took place, a new contract had just been concluded, which lasts until July, 1997. Many members, but not all, considered it a reasonable and satisfactory settlement. The grievance system documented in the contract is simple. In response to malpractice or misconduct, a voluntary consultation will be held where two supervisors, a worker, and a steward will go through the incident. If nothing is settled, the case will become formal. Then, there will be a one-session discussion between the union and labour officials before they go to arbitration. Arbitrators have to be chosen from a list in the collective agreement, and no legal counsels are allowed in arbitrations. Except for a few issues, such as management policies, expedited arbitration can be used if both parties agree to use it. Mediation and other alternative dispute resolution procedures are not documented in the collective agreement or mentioned by the labour officers. Procedures 70 Studying ethnic groups in a multicultural society is similar to studying widely varying and geographically dispersed cultural groups (Berry et al., 1992: 292). Making correct inferences on the influence of ethnicity is easily plagued by unforeseen problems in sampling, procedures of experiments, instruments, and even translation (Lonner & Berry, 1986). Even conducting the study with great care, researchers may still find competing explanations for their results after finishing the study. Although this consequence is true for any scientific endeavour, I believe that it is more common in cross-cultural and ethnic research. One way to deal with competing explanations is to collect broad, contextual data in addition to quantitative, specific data, so that other researchers can assess the assumptions and judgment calls made in testing specific hypotheses (Lonner & Berry, 1986: 90). Accordingly, data of this study came from three sources: review of documents and books about the company, interviews with insiders of the organization, and a survey on the workers of the organization. Documents. Annual reports of the corporation, collective agreements, and two books are reviewed. One book revealed the unfair treatment of women in the corporation (1990), and the other attacked the stress the corporation cast on its outdoor workers in Edmonton (1986). Both books, however, were written in the 80's and fail to capture the recent "defrosting" of the union-management relationship in the corporation. Interviews and written comments. The purpose of interviews is to obtain the perspectives of insiders and to avoid omitting major relationships in the "nomological net." To this end I conducted semi-structured interviews with two supervisors and two union stewards assigned by the management and union. Al l of the interviewees have served the 71 organization for at least 15 years, and are familiar with the technical and social aspects of the organization. Appendix A presents the interview outline. The length of interviews varied from 40 minutes to 90 minutes. Another source of interview information came when I was distributing questionnaires at the plant. Some staff were interested in talking to me about their workplace. While these interviews were not structured, they supplied additional information on the perspectives of workers. A third source of information was comments on returned questionnaires and attachments (three of them are more than 3-pages long). Information from these interviews was used in describing the organization above. It will also be used when discussing the implications of survey results in Chapter 6. Survey. The survey employed only the in-house staff who performed the core tasks of the company. Outdoor and indoor maintenance staff were excluded. As part-timers are only called in when work requires, the exact numbers of them are unknown when I distributed the questionnaires. However, according to a senior labour officer, full-time workers were 650 and part-time workers were about 300. The questionnaire was handed out inside the plant in two ways: 384 were handed out when staff were having meal or coffee breaks at the cafeteria; another 266 were given out at the main gate when the staff were leaving or coming to work. Each staff received a package composed of a cover letter, a questionnaire, a slip, and a prepaid, self-addressed envelope. The cover letter explained the purpose of the study and stated the rights of potential respondents (see Appendix B). It was signed by me (the researcher), a senior labour officer, and the president of the union local. The slip was for 72 people who did not want to fill out the questionnaire. It asked their age, gender, reason for not responding, and whether they had filed a grievance last year. Mitchell (1985) recommends this way to get some information about non-respondents. Unfortunately, most people did not bother to mail the slip when they decided not to respond. There were two versions of questionnaires (described in detail below). One was for testing the hypotheses and the other was for ascertaining the validity of two important constructs - ethnic belonging and individualism-collectivism. Although the two constructs were measured with validated scales, confirming their validity is recommended when they are used for different samples and in new settings (Mitchell, 1985). Length rendered it impossible to include items used for validating constructs and testing hypotheses in one questionnaire. Doing so would make the questionnaire over 5-pages long, and response rate might drop to an unacceptable level. The questions were written in simple English. A labour officer and two union officials first examined the questionnaires and suggested a few changes on wordings. The revised questionnaires were reviewed by shop stewards and supervisors for comments during interviews. Al l of them thought that the staff, even non-native English speakers, could understand the questions, and that the questions were interesting and comprehensive. 139 of the questionnaires were received with a return rate of 21%. Among the returned, nine were discarded because three were filled in by wrong respondents, and six of them were only slips for respondents who did not want to fill in the questionnaire. Among the 130 usable questionnaires, 108 of them are for hypothesis-testing, and 22 of them are for construct validation. 73 Questionnaire for Hypothesis Testing The questionnaire measured five groups of variables: grievance behaviours, three components related to the multiculturalism framework, additional variables for the framework, and demographic variables (see Appendix C). Table 2 presents the items and descriptive statistics of the scales. Dependent variables. Respondents first gave the number of grievances filed over last year, the types of the grievances, and a description of the grievances. They then checked statements concerning their attitudes and beliefs on five-point scales (1-Strongly disagree; 5-strongly agree), except stated otherwise. Three statements were used to measure job satisfaction: 1. The performance of management is satisfactory. 2. In general, I like working here. 3. Management do not treat me fairly. (-) Adapted from Bemmels et al. (1991), three statements were used to measure complaint behaviours: 1. I often tell supervisors about my unhappiness of the workplace. 2. I voice my work dissatisfactions to shop stewards. 3. I discuss my unhappiness at work with coworkers of my ethnic origin. Two statements on informal settlement between shop stewards and supervisors were also adapted from Bemmels et al. (1991): 1. Union stewards discuss my complaints through informal discussion with my supervisor. 2. Union stewards try settling my potential grievance with management before they file a 74 Table 2 : Measurement of Variables (Hypothesis-Testing) Variable Measurement S (sd) Grievance Behaviours No. of Grievances How many grievances have you f i l e d l a s t year? Complaint Supervisor Shop Steward Coworkers of same Ethnic group Job S a t i s f a c t i o n (3 items: a = .53) Informal Settlement by Shop Stewards (2 items: r = .50) I often t e l l supervisors about my unhappiness of the workplace I voice my work d i s s a t i s f a c t i o n to shop stewards I discuss my unhappiness at work with coworkers of my ethnic o r i g i n a. The performance of management i s s a t i s f a c t o r y b. In general, I l i k e working here c. Management do not t r e a t me f a i r l y (-) a. Stewards discuss my complaints through informal discussion with my supervisors b. Stewards t r y s e t t l i n g my p o t e n t i a l grievances with management before they f i l e a grievance Ethnic Belonging (3 items: a = .71) Three Components of the M u l t i c u l t u r a l i s m Framework a Atmosphere of M u l t i c u l t u r a l i s m (2 items: r = .33) Engl i s h A b i l i t y I am happy that I am a member of the ethnic group I belong to b. I do not f e e l good about my c u l t u r a l and ethnic background (-) c. I have a strong sense of belonging to my own ethnic group a. People, regardless of t h e i r ethnic backgrounds, a l l support me i n the workplace b. C u l t u r a l d i v e r s i t y i s not accepted i n the workplace (-) a. I speak well i n English (2 items: r = .77) b. I write well i n English Individualism-Collectivism Value Coworker C o l l e c t i v i s m (4 items: oC = a. One needs to return a favor i f a coworker leads a helping hand .44) b. Assistance from coworkers i s necessary for doing a good job c. I l i k e to l i v e close to my good friends d. If coworkers group together to help each other, they can gain a l o t at no cost 1.3(2.7) 2.8(1.1) 3.3(1.0) 2.9(1.2) 9.1(2.0) 6.2(1.6) 9.6(1.7) 7.2(1.6) 6.8(1.2] 14.3(2.1) 75 Table 2 (Continued): Measurement of Variables (Hypothesis-Testing) Variable Measurement 2 (sd) Individualism-Collectivism Value General C o l l e c t i v i s m (3 items: a = .32) a. I know how to make friends with my 11.7(1.8) neighbour b. Even i f a c h i l d wins a pr e s t i g i o u s p r i z e , the parents should not f e e l honoured i n any way (-) c. I would not l e t my parents use my car ( i f I have one), whether they are good d r i v e r s or not (-) Ethnic Status Ethnic Status (2 items: r = .77) a. My ethnic group has a high status i n society b. People of other ethnic groups never discriminate against my own ethnic group Evaluations of the Grievance Procedure 5.5(1.4) Importance of Grievance Procedure The grievance procedure i s important f o r the workplace 4.3( .9) Fairness of Grievance Procedure (3 items: a = .72) Instrumentality of Grievance Procedure (2 items: r = .48) Producing Grudges A f t e r Grievance F i l i n g (2 items: r = .59) Worry of R e p r i s a l from Management (2 items: r = .13) a. The grievance procedure produces f a i r 9.3(2.2) outcomes b. The grievance procedure i s designed i n a f a i r way c. Workers receive f a i r treatment from management while t h e i r grievances are being processed a. Those who f i l e grievances usually get 6.1(1.4) favourable outcomes b. Workers usually gain good outcomes from grievance decisions a. F i l i n g a grievance produces bad f e e l i n g s among the disputing p a r t i e s b. F i l i n g a grievance damages the r e l a t i o n s h i p between management and workers 6.3(1.9) a. Management w i l l t r y to r e t a l i a t e i f a 6.4(1.5) worker f i l e s a grievance b. Management does not harass workers a f t e r they f i l e d a grievance 76 Table 2 (Continued): Measurement of Variables (Hypothesis-Testing) Variable Measurement S (sd) Demographic Composition Variables Proportion of In my workplace, there are only a few minority 3.1(1.2) Minority Managers managers and supervisors Proportion of In my workplace, there are a large number of 3.8( .8) Minority Coworkers minority coworkers Proportion of Most of the union stewards who serve me at the 3.4(1.2) Stewards Who Speak workplace speak my f i r s t language F i r s t Language of Respondents [5-pt.] Proportion of Few of my supervisors are of my ethnic Supervisors Who background 3.0(1.1) Have Same Ethnic Background [5-pt.] 77 grievance. Three components of the multiculturalism framework. Two items were used to measure English ability (1-strongly disagree; 4-strongly agree): 1. I speak well in English. 2. I write well in English. Based on a comprehensive review on the concept of ethnic identity (Phinney, 1990), Phinney (1992) proposed and tested The Multigroup Ethnic Identity Measure. This Measure is designed for gauging ethnic identification for a variety of ethnic groups. The Affirmation and Belonging subscale rather than the full scale is used because this subscale measures, as expected above, the symbolic ethnic identification of individuals. The Cronbach a of the ethnic belonging subscale for High school students is .75 and that of college students is .86. The validity and reliability of the scale, I believe, are likely to be applicable to adults, as these items were generated based on studies using adult subjects (Phinney, 1990).2 Respondents will first indicate their self-perceived ethnic group on an open-ended question, "Please fill in: In terms of ethnic group, I consider myself to be ." They will then respond to three items on 4-point scales (1-strongly disagree; 4-strongly agree): 1. I am happy that I am a member of the ethnic group I belong to. 2. I have a strong sense of belonging to my own ethnic group. 3. I do not feel good about my cultural or ethnic background. (-) 2I prefer the scale of Phinney to the Group Identification Scale (Hinkle et al., 1989; Oaker & Brown, 1986), developed along the line of social identity theory, because the latter has never been applied to measure ethnic identity and have been restricted in studying monocultural work groups and small group research. 78 The two defining aspects of the atmosphere of multiculturalism were measured by two items modified from Berry et al. (1992: 287): 1. People, regardless of their ethnic background, all support me in the workplace. 2. Cultural diversity is not accepted in this organization. (-) Additional constructs for the multiculturalism framework. Two items measure ethnic status: 1. My ethnic group has a high status in society. 2. People of other ethnic groups never discriminate against my own ethnic group. An 8-item scale was proposed by Triandis (1988: 90) to measure individualism as a personality variable. Alternatively, Hui and his colleagues (e.g. Hui, 1988; Hui & Yee, 1994) provided a cross-culturally validated instrument, INDCOL. INDCOL was composed of 63 items (a = .74; Triandis et al., 1985: 406), corresponding to six subscales: spouse, parent, kin, neighbour, friend, and coworker. Six subscales were developed, for collectivistic disposition varied across target persons (Hui & Triandis, 1986). Using U.S. and Hong Kong university students as subjects (Hui, 1988), the subscales had as ranging from .41 (spouse) to .76 (parent). Hui and Yee (1994) used Hong Kong managers as subjects and found shortened subscales to have as ranging from .38 (parent-spouse) to .73 (neighbour). For the purpose of this study, I employed a shortened coworker INDCOL subscale and constructed a scale to measure general I-C value. The coworker I-C subscale consisted of five items that have high factor loadings (Hui & Yee, 1994):3 3This measurement is preferred to other possible scales. The scale developed in Wagner and Moch (1986) was not for ethnic or cross-cultural research; that of Earley (1993) was emphasized on goal and performance in the workplace, and that of Bochner and Hesketh 79 1. One needs to return a favor if a coworker lends a helping hand. 2. Assistance from coworkers is necessary for doing a good job. 3. I would help a coworker who needs money to pay utility bills. 4. If coworkers group together to help each other, they can gain a lot at no cost. 5. I like to live close to my good friends. Two other items of the subscale were dropped as they were too long and difficult to be rephrased for this study. The general individualism scale was formed specifically for this study. It is composed of four items which had the highest loadings on the other four I-C subscales, which focuses on different target persons (Hui and Yee, 1994). As a result of its composition, this scale measures the general I-C value across a variety of different target persons, and I do not expect it to have a high a: 1. I know how to make friends with my neighbours. 2. How people spend their income is not the business of their relatives (cousins, uncles). (-) 3. I would not let my parents use my car (if I have one), whether they are good drivers or not. (-) 4. Even if a child wins a prestigious prize, the parents should not feel honoured in any way. (-) Evaluations of the grievance procedure. People's evaluations of the grievance process were measured with different sets of items. A statement assessed the importance of the grievance procedure: (1994) lacks the validation vigour of INDCOL. 80 1. The grievance procedure is important for the workplace. The instrumentality of the procedures to workers was measured by two items: 1. Those who filed grievances usually get favourable outcomes. 2. Workers usually gain good outcomes from grievance decisions. The perceived fairness of the procedure was measured by three items to reflect the distributive, procedural, and interactional aspects of fairness (Tylor & Bies, 1990): 1. The grievance procedure produces fair outcomes. 2. The grievance procedure is designed in a fair way. 3. Workers receive fair treatment from management while their grievances are being processed. Feeling of animosity between workers and management after filing grievances was measured by two items: 1. Filing a grievance produces bad feelings among the disputing parties. 2. Filing a grievance damages the relationship between management and workers. Worry of reprisal by management was assessed by two items: 1. Management will try to retaliate if a worker files a grievance. 2. Management does not harass workers after they filed a grievance. (-) Demographic composition variables. Different from studies on the effects of demographic compositions, this study asked subjects' perceptions of proportions of minorities and majorities instead of estimating the proportions using objective data (e.g., Tsui & O'Reily, 1989; Jackson, 1992). Four items asked the majority and minority compositions of coworkers, supervisors, and stewards in the workplace. 81 1. In my workplace, there are a large number of minority coworkers. 2. In my workplace, there are only a few minority managers and supervisors. 3. Most of the union stewards who serve me in the workplace speak my first language. 4. Few of my supervisors are of my ethnic background. Demographic variables. Several demographics information were collected: age, sex, years stayed in Canada, length of service with the present employer, years of schooling, job nature, position, contract type, and whether one worked as a shop stewards over last year and before. The first language of respondents was also asked to help define the ethnic group one belonged to (Phinney, 1992). Questionnaire on Convergent and Divergent Validity This version of the questionnaire was used to ascertain the psychometric properties of the two major constructs - I-C value and ethnic belonging (see Appendix D). Specifically, I tried to demonstrate their convergent and discriminant validity by showing that these constructs were related, with theoretical reasons, only to specific constructs and not others. Some items are present in this questionnaire to make sure that both versions look alike and are relevant to grievance behaviours. These items include those about grievance behaviours, evaluations of the grievance procedure, and demographics. Other items except those for the two central constructs are absent from this version. The following presents scales in this version that are not included in the hypothesis-testing version (see Table 3 for the items and descriptive statistics). Individualism-Collectivism value. Items for coworker and general I-C value are included. According to Triandis et al. (1985: 410), general collectivism are expected to be 82 Table 3: Measurement of Variables (Construct Validation) Variable Measurement 2 (sd) Ethnic Behavior (2 items: r Variables for Testing Ethnic I d e n t i f i c a t i o n a. I am active i n organizations or s o c i a l ,64) groups that include mostly members of my own ethnic group b. I p a r t i c i p a t e i n c u l t u r a l p r a c t i c e s of my own ethnic group, such as s p e c i a l food, music or customs 5.4(1.9) Other-Group Orientation (3 items: a = .81) b. I l i k e meeting and getting to know people from ethnic groups other than my own I sometimes f e e l that i t would be better i f d i f f e r e n t ethnic group didn't t r y to mix together I don't t r y to become friends with people from other than my own Variables f o r Testing Individualism-Collectivism 12.6(2.1) Job Involvement (3 items: a = .85] The most important things that happen to me involve my present job Most of my i n t e r e s t s are centred around my job I consider my job to be very c e n t r a l to my l i f e 8.0(2.9) S o c i a l Support (6 items: a = .80) Machiavellianism (2 items: r = .53) a. I receive useful information from a l o t 17.5(4.1) of people b. Many people around me provide me with encouragement c. Almost everyone says things that r a i s e my self-confidence d. Many people l i s t e n to me when I need to t a l k e. People around me show that they care about me as a person f. My friends and my family understand the way I t a l k and f e e l about things a. Anyone who completely t r u s t s anyone e l s e i s 6.2( .8) asking for trouble b. It i s hard to get ahead without c u t t i n g corners here and there 83 positively related to social support received by the respondents (r from .29 to .42) because cohesive and trustworthy ingroups provide quality social support by allowing the person to express emotions, seek advice, and solicit encouragement. This relationship may exist and is strong only for general I-C value but not coworker I-C value because general I-C measures how cohesive the person is with many of the ingroups around him or her whereas coworker I-C measures only that with his or her coworkers. Six items were adapted from Vinokur, Schul, and Caplan (1987) to measure social support (a ranges from .81 to .85). These items assess perceived quality of social support from people around the respondents. They measured the three functions of social support, including emotional, appraisal, and informational support. These items are as follows: 1. Many people around me provide me with encouragement. 2. I receive useful information from a lot of people. 3. Almost everyone says things that raise my self-confidence. 4. People around me show that they care about me as person. 5. Many people listen to me when I need to talk. 6. My friends and my family understand the way I think and feel about things. In addition, Triandis et al. (1985) found that the two I-C value constructs are negatively related to alienation , conceptualized as a state of loss of individuality (Middleton, 1962). Using Middleton's scale, Triandis et al. (1985) found that scores of collectivism correlated with alienation scores (r = -.27). As this study is in work setting, I chose to use the Job Involvement Scale (a = .87; test-retest reliability = .85) derived by Kanungo (1982) after his careful study of work alienation. Based on Triandis et al., I expect a moderately but 84 positive relationship between job involvement and coworker collectivism. General I-C value should show null or weaker relationships because it is not specific about the workplace. Three items were adapted from Kanungo's scale: 1. The most important things happen to me involve my present job. 2. Most of my interests are centred around my job. 3. I consider my job to be very central to my life. While I-C value is expected to be related to the social support and job involvement in specific ways, no relationship is expected for Machiavellianism personality (Triandis et al., 1985). If general and coworker I-C were not related to Machiavellianism, it suggested that the scales were not measuring a personality construct that had already been well-studied. This gives some support, though not complete, to the argument that the I-C scales measure some thing other than Mach and, thus, the discriminant validity of the constructs. Three items that show the best discrimination power were adapted from the Mach scale (Christie & Geis, 1970: 17): 1. The best way to handle people is to tell them what they want to hear. 2. Anyone who completely trusts anyone else is asking for trouble. 3. It is hard to get ahead without cutting corners here and there. Ethnic belonging. The validity of ethnic belonging was assessed similar to the above. I used the Ethnic Behavior subscale of the Multigroup Ethnic Identification Measure from Phinney (1992) to assess convergent validity.4 This subscale is expected to be related to the ethnic belonging subscale used in the hypothesis testing questionnaire. The items of the 4He gave no reliability index, but stated that the items were of good quality. 85 Ethnic Behavior subscales are as follows: 1. I am active in organizations or social groups that include mostly members of my own ethnic group. 2. I participate in cultural practices of my own ethnic group, such as special food, music or customs. On the other hand, discriminant validity was assessed by attitudes toward outgroups (a = .71 for high schoolers; a = .74 for college students) which were not related to any of the two subscales. Three items were used to measure this construct: 1. I like meeting and getting to know people from ethnic groups other than my own. 2. I don't try to become friends with people from other ethnic groups. (-) 3. I sometimes feel that it would be better if different ethnic groups didn't try to mix together. (-) 86 Chapter 5: Results In the following I will first present results concerning the validation of constructs. This will be followed by a discussion on data concerning self-identified ethnic groups. Two different categorizations of ethnic groups and the characteristics of these groups will be discussed. Lastly, I will present regression results concerning the hypotheses of the study. This discussion starts with the three components of the multiculturalism framework, moves on to grievance behaviours, and finishes with the evaluations of the grievance procedure. Rejected and supported hypotheses are indicated in Table 1, which is reproduced on the next page. Validation of Constructs Ethnic belonging. Al l measures have reasonable reliability. In particular, the as of the ethnic belonging scale are .73 for the hypothesis-testing sample, .60 for the construct-validation sample, and .71 for the pooled sample. The pattern of correlations between the three focal measures - ethnic belonging, ethnic behaviour, and other-group orientation -resembles the findings of Phinney (1992: 167), although the magnitude of the correlations differs. Ethnic belonging is related to ethnic behaviour (r = .33 vs. .46), though insignificantly (p_ < .13), and is weakly related to other-group orientation (r = .10 vs. .06). In addition, ethnic behaviour is negatively and significantly related to other-group orientation (r = -.45, p_ < .05, vs. -.17). Together, these results support the validity of the ethnic belonging scale adapted from Phinney (1992) for this study. Individualism-collectivism (I-C). The coworker I-C and general I-C scales (r = . 19) do not have the decent as as does the ethnic belonging scale. Although the coworker I-C 8 7 Table 1: Hypotheses of the Study Grievance Behaviours Hla: Job dissatisfaction is positively related to the frequency of complaining to supervisors. [/] Hlb: Job dissatisfaction is positively related to the frequency of complaining to shop stewards. [S] Hlc: Job dissatisfaction is positively related to the frequency of complaining to coworkers of own ethnic background. [/] Hid: Job dissatisfaction is positively related to the number of grievances filed. [/] H2a: Complaining to supervisors is positively related to the number of grievances filed. H2b: Complaining to shop stewards is positively related to the number of grievances filed. H2c: Complaining to coworkers of own ethnic background is negatively related to the number of grievances filed. [/] H3: Informal settlement by shop stewards is negatively related to the number of grievances filed. Linguistic Vitality Hypothesis H4: English ability is associated with ethnic belonging, negatively for the minorities but positively for the majorities. [P] Linguistic Sharing Hypothesis H5: English ability is positively related to perceived atmosphere of multiculturalism. Linguistic Accommodation Hypothesis H6a: English ability is positively related to job satisfaction. H6b: English ability is positively related to complaints to supervisors. [</] H6c: English ability is positively related to complaints to shop stewards. H6d: English ability is not related to complaints to coworkers of same ethnic background. [•] H6e: English ability is positively related to the number of grievances filed. Note. [>/] indicates that the hypothesis is supported. [P] indicates that the hypothesis is partially supported. No sign indicates that the hypothesis is not supported. 8 8 Table 1 (Continued): Hypotheses for the Study Ethnic Confidence Hypothesis H7: Ethnic belonging is related positively to perceived atmosphere of multiculturalism for the minorities, but negatively to perceived atmosphere of multiculturalism for the majorities. [P] Primary Multiculturalism Hypothesis H8a: Perceived atmosphere of multiculturalism is positively related to job satisfaction. H8b: Perceived atmosphere of multiculturalism is negatively related to complaints to supervisors. H8c: Perceived atmosphere of multiculturalism is negatively related to complaints to shop stewards. H8d: Perceived atmosphere of multiculturalism is negatively related to complaints to coworkers of own ethnic background. H8e: Perceived atmosphere of multiculturalism is negatively related to the number of grievances filed. Secondary Multiculturalism Hypothesis H9a: When perceived atmosphere of multiculturalism is high, ethnic belonging is positively related to job satisfaction. [/] H9b: When perceived atmosphere of multiculturalism is high, ethnic belonging is negatively related to complaints to supervisors. [/] H9c: When perceived atmosphere of multiculturalism is high, ethnic belonging is negatively related to complaints to shop stewards. H9d: When perceived atmosphere of multiculturalism is high, ethnic belonging is negatively related to complaints to coworkers of own ethnic background. H9e: When perceived atmosphere of multiculturalism is high, ethnic belonging is negatively related to the number of grievances filed. HlOa: When perceived atmosphere of multiculturalism is low, ethnic belonging is negatively related to job satisfaction. HlOb: When perceived atmosphere of multiculturalism is low, ethnic belonging is positively related to complaints to supervisors. HlOc: When perceived atmosphere of multiculturalism is low, ethnic belonging is positively related to complaints to shop stewards. HlOd: When perceived atmosphere of multiculturalism is low, ethnic belonging is positively related to complaints to coworkers of own ethnic background. HlOe: When perceived atmosphere of multiculturalism is low, ethnic belonging is positively related to the number of grievances filed, [y] Note. [>/] indicates that the hypothesis is supported. [P] indicates that the hypothesis is partially supported. Otherwise, the hypothesis is not supported. 89 Table 1 (Continued): Hypotheses for the Study Ethnic Status Hypothesis H l l : Ethnic status is positively associated with ethnic belonging. H12: Ethnic status is positively associated with perceived atmosphere of multiculturalism. [/] Individualism-Collectivism Value Hypothesis H13: Collectivism is positively related to ethnic belonging. [/] H14: Collectivism is positively related to perceived atmosphere of multiculturalism. [/] H15: Collectivism is negatively related to job satisfaction. [/] H16a: General collectivism is negatively related to the frequency of complaining to supervisors. H16b: General collectivism is negatively related to the frequency of complaining to stewards. [/I H16c: General collectivism is negatively related to the number of grievances filed. [/] H17a: Coworker collectivism is positively related to the frequency of complaining to supervisors. [/] H17b: Coworker collectivism is positively related to the frequency of complaining to stewards. H17c: Coworker collectivism is positively related to the number of grievances filed. [>/] H18: Collectivism is positively related to the frequency of complaining to coworkers of same ethnic background, [/\ H19a: General collectivism is negatively related to perceived instrumentality of grievance procedures. H19b: Coworker collectivism is positively related to perceived instrumentality of grievance procedures. H20a: General collectivism is positively related to perceived animosity after grievance filing. M H20b: Coworker collectivism is negatively related to perceived animosity after grievance filing. Note. [•] indicates that the hypothesis is supported. [P] indicates that the hypothesis is partially supported. Otherwise, the hypothesis is not supported. 90 scale has an a of .66 for the construct-validation sample, the a for the hypothesis-testing sample is .36, and that for the pooled sample is .44. In comparison, the a's in Hui's studies (Hui, 1988; Hui and Yee, 1994) range from .52 to .58. The item, "I would help a coworker who needs money to pay utility bills," was dropped because it correlated negatively with the aggregate score of the other four items. Similar results are found for the general I-C scale. The item, "How people spend their income is not the business of their relatives (cousins, uncles)," was dropped due to the negative correlation with the aggregate score. The resultant as are .30 for the construct-validation sample, .28 for the hypothesis-testing sample, and .32 for the combined sample. Weak reliability is, more or less, expected given that the scale was composed of items from three sub-scales of the INDCOL. The idea was to aggregate these items which tapped the I-C value of different target persons, so that the general manifestation of I-C value could be measured. To cover appropriate breath of general I-C across target persons, only one out of five or six items was taken from each subscale. Since the items were heterogeneous, the low reliability might not be caused by unreliable items. As stated by Humphrey (1949), "one should not always try to maximize test reliability since it is a function of homogeneity as well as item reliability." Even though the reliability is low, the scales are not necessarily useless. Indeed, Cronbach (1951) showed that a general factor can be well measured by aggregating items of which systematic variances consisted of as little as 9% of the variance of the general factor. In his example, the intercorrelations between the items are between .09 to .18. Yet, when the number of items were three, they tapped about 20% of the true variance of the general factor. Therefore, although low reliability entails cautious interpretations of 91 results, it is possible that the I-C scales in this study have tapped a significant portion of true variance. Nonetheless, as low as indicate that the items tap only a fraction of the systematic variance of the construct, error variance which is random and not related to any constructs may blur true relationships between I-C value and other constructs. We should, therefore, be especially careful in cases where null relationships are not rejected. Bearing this caution in mind, the scales seems to be valid, for the correlations between the I-C scores and other measures corroborated most of the a priori expectations about the convergent and discriminant validity of the scales. Discriminant validity is supported as the correlations between the two I-C scores and an established personality construct - Machiavellianism - are low and insignificant (r = .18 & .22). This means that the scales are not a replicate of a well-founded construct. However, null results may reinforce the argument that the scores are random noise if no other significant correlations are found, especially those expected a priori. Further analysis shows that this is not the case. As expected, the coworker I-C scores are related to work involvement scores (r = .37, p_ < .10), but not social support scores (r = -.10, p_ < .10). The coworker scores are related to work involvement scores because both of them are situation-specific (Fishbein & Ajzen, 1975). However, social support is related to neither the general I-C scores (r = -.24, n.s.), nor coworker I-C scores (r = -.07). Two possible reasons may explain this finding, apart from concluding that the scores are merely noise. One plausible reason is that the null effect may be caused by ceiling effect. As most subjects scored high on the three five-point 92 scales (E = 12.0; sd = 1.7), the scale may be cut off at the high end, and results in a truncated distribution and weak correlation. However, plotting the scores does not reveal a truncated distribution. Another plausible explanation is that negative and positive relationships exist for different ethnic groups. White Canadians are relatively more individualistic (M — 11.7; sd = 1.7; N = 14) than other ethnic groups (M = 12.7; sd = 1.0; N = 7). Triandis et al. (1985: 412) found that the relationship between social support and I-C was in fact curvilinear. The relationship was negative for people high in individualism, but was positive for people high in collectivism. Perhaps, a null result was found in this study because positive relationship in minorities was cancelled by negative relationship in whites. Further analysis found that this explanation was possible, as social support scores correlated negatively with I-C scores for whites (r = -.38) but positively for other ethnic groups (r = .23). In hindsight, whites who behave similar to collectivists may find it hard to get along with other whites who tend to behave like individualists, and, hence, behaving collectively may make whites receive less quality social support, as indicated by the negative correlation above. Al l in all, the validity of the I-C scales receives some support, though it is not as strong as that of the ethnic belonging scale. Other constructs. The measurement of other constructs is straight-forward. As shown in Table 2, the scales have either reasonable reliability or highly correlated items. The only exception is worry of reprisal from management after filing grievances. The scores of the two items have low correlation (r = .13, n.s.). The reason may be that retaliation, measured in one item, implies a serious, noticeable incident, and does not happen as often as 93 harassment, measured in the other item. Subsequent analyses, however, found that results were similar for scores of individual items and for aggregate scores. For simplicity, only the aggregate scores were used in the final analysis. Common method variance. Since this study collected data through a single, self-administered questionnaire, several reasons might cause the observed correlations between the variables rather than genuine relationships (Bemmels, 1994; Podsakoff & Organ, 1986; Kemery & Dunlap, 1986). First, because the measures came from one source, any systematic defects in the questionnaire might result in significant correlations. The second reason is the consistency motif. When asked to report to a series of related events or evaluations, people may strive to report a consistent line of answers based on their implicit theory about the questioned events. Third, the situational cues may be common to respondents, leading them to give similar answers and causing artificial covariation. Based on Podsakoff and Organ (1986), two tests were made on the 33 items in section 2 of the questionnaire (Appendix C) to assess common method bias.1 These items comprise all the socio-psychological variables, except ethnic belonging and English ability. They are of particular interest because all of them were measured in the same format (1-disagree, 5-agree), and they were used to measure both the dependent and independent variables of the study. The first test was a principal component analysis with no rotation on the 33 items. If a large amount of common method variance existed, a general factor should be extracted as JThe original questionnaire has 35 items, but one item for coworker I-C and one for general I-C scales were deleted. 94 the first factor. The analysis resulted in eleven factors with eigenvalues greater than 1. A scree test revealed a 3-factor model. The eigenvalues of the first six factors were respectively 4.1 (13%), 3.6 (11%), 2.5 (7.4%), 2.1 (6.4%), 2.0 (6.2%), and 2.0 (6.0%). As the eigenvalues plateau after the third factor, a 3-factor solution is indicated, and the first factor does not explain substantially more total variance than the second factor. The second test was based on the argument that, if common method bias was serious, a one-factor solution should be the case. Neither the unrotated solution nor a rotated solution using varimax method suggested that this was true. As shown above, the first and second factors of the unrotated model explained similar levels of variance, and just like the unrotated solution, the first and second factors of the rotated solution explained 12.5% and 11.0% of the total variance. Not all the items load on the first factor, as common method variance would suggest. For the unrotated solution, the first factor had moderate to high loadings on eight of the items (.37 to .75), but not on others (< .30). For the rotated solution, it loaded on four of the items (.34 to .86) and not on others. Therefore, because a one-factor model and a large number of loaded items on the first factor were not in evidence, common method variance does not seem to be serious in this study. Ethnic Categories Respondents were asked to name the ethnic group of which they considered themselves to be a member. There was a wide variety of responses. For the purpose of this study, individuals had to be classified into broader and meaningful categories, so that analysis became possible. After trying a few categorizations, taking into consideration the practical and theoretical significance, I decided to report the results of two different categorizations. 95 One classifies respondents into two groups, visible ethnic minorities and majorities. The other classifies them into three categories: white Canadians, Canadians of European descent other than British, and visible minorities. The last category is the same as the minority category in the two-group classification. Three-group categorization. As shown in Table 4, most people identified themselves as Canadians, whites, WASP, Caucasians, and English Canadians (Group I). People who reported these labels were clustered together because they represent the white Canadians as a group. Others identified themselves as Canadians of a European descent, such as Italian Canadians and Danish Canadians, or simply the natives of a European country, such as German and Ukrainian (Group II). They were put into another group because reporting either a European or a hyphenated Canadian labels usually suggest that they were not born in Canada or were the second generation of immigrants from Europe (Berry et al., 1977). Group III were people who identified themselves as a visible minorities in Canada, such as Filipino, and Chinese. Hyphenated Canadians who reported a visible-minority label, such as Indo-Canadians, were also put into this group. The first language of respondents were used to assist the above classification. Appendix E lists all group labels, first languages, and the corresponding numbers of self-identified respondents. Tables 4 and 5 present the characteristics of these three groups. They are similar in several demographic attributes, including age, education, and years served in the organization. The three groups also reported similar levels of perceived multicultural atmosphere, ethnic belonging, ethnic status, proportion of minority workers, and coworker I-C and general I-C values. They differ in several ways, however. First, their years of living 96 in other countries are different. The lengths of Groups I and II are similar, whereas that of Group III is remarkably longer. Moreover, reported English ability also differs in similar way; Groups I and II reported having ability better than that of Group III. In addition, they saw different proportions of minority stewards, supervisors, and managers in the workplace. Respondents in Group III saw fewer minorities in these positions, whereas those in Categories I and II saw more of them. These differences may result from having different definitions of "many" and different personal networks in the workplace. Lastly, Group III complained more to their ethnic coworkers about job dissatisfaction than Groups I and II, which were similar in this aspect. Two-group categorization. As Groups I and II were similar to each other but different from Group III in many aspects, it was deemed useful to pool Groups I and II together. This classification creates a contrast between visible minorities and other Canadians. This contrasthas been a concern in the 90s. For instance, when comparing the ethnic relations in the U.S. and Canada, Reitz and Breton (1994) focused on studying visible minorities in contrast to other ethnic groups because these immigrants have replaced Europeans as the major source to North America. Tables 4 and 5 present the characteristics of the Minority and Majority groups. The two groups were similar in several demographic attributes, including age, education, and years served. They also reported similar levels of ethnic belonging, ethnic status, perceived proportion of minority workers, coworker I-C scores, and general I-C scores. However, they are different in several aspects. First, the first languages of those in the minority category are foreign, chiefly Asian, such as Hindi, Cantonese, and Tagalog, 97 Table 4: Characteristics of Different Ethnic Groups Group I Group II Group III Variable Overall Canadian European Canadian Visible Minority Majority Minorit N 108 56 13 39 69 39 Age [Yrs] 43.2 ( 9.3) 43.5 (10.1) 42.1 (11.9) 43.2 ( 9-1) 43.2 (10.4) 43 . 2 ( 7.2) % of Women 49 42 62 54 46 54 Working as a Steward over Last Year (number) 4 4 0 0 4 0 Education [Yrs] 14.2 ( 3.4) 14.4 ( 4.1) 13.9 ( 2.1) 14. 1 ( 2.5) 14.3 ( 3.8) 14. 1 ( 2.5) Years Lived in Other Countries [Yrs] 8.7 2.3 3.9 18.8 2 . 6 18 . 8 Years Lived in Other Countries [Yrs] (10.9) ( 6.7) ( 8.5) ( 8.2) ( 7.0) ( 8.2) Length of Service in Company [Yrs] 14.1 ( 7.9) 14. 0 ( 8.7) 12.4 ( 6.4) 14.8 ( 7.2) 13 .7 ( 8.3) 14.8 ( 7.2) English Ability [8-pt.] 6.8 7.1 7.0 6.3 7.1 6.3 English Ability [8-pt.] ( 1.2) ( 1.1) ( 1.5) ( 1.2) ( 1.2) ( 1-2) Positive Atmosphere of Multiculturalism [10-pt.] 7 . 2 7.5 7.5 6.8 7.5 6.8 Positive Atmosphere of Multiculturalism [10-pt.] ( 1-6) ( 1-4) ( 1-6) ( 1.7) ( 1.5) ( 1.7) Ethnic Belonging [12-pt.] 9.6 ( 1-9) 9.7 ( 1.8) 9.8 ( 1.6) 9.2 ( 2.2) 9.7 ( 1-7) 9.2 ( 2.2) Note. Figures in brackets are sds. Rows of figures underlined are different at p_ < Table 4 (Continued): Characteristics of Different Ethnic Categories Group I Group II Group III European Visible Variable Overall Canadian Canadian Minority Majority Minority Positive Ethnic Status [10-pt.] 5.5 ( 1.4) 5.4 ( 1.3) 5.9 ( 1-8) 5.4 ( 1.3) 5.5 ( 1-4) 5.4 ( 1.3) Coworker Subscale of Collectivism [20-pt.] 14 . 3 ( 2.2) 14.3 ( 2.3) 13.6 ( 2.3) 14.6 ( 2.0) 14. 1 ( 2.3) 14.6 ( 2.0) General Collectivism [15-pt.] 11.7 ( 1.8) 11. 6 ( 1.6) 11. 3 ( 2.5) 11.9 ( 1.7) 11. 5 ( 1-8) 11.9 ( 1-7) Many Minority Coworkers [5-pt.] 3 . 8 ( 1-0) 3.8 ( 1.0) 3.8 ( 1-2) 3.9 ( -8) 3.8 ( 1.1) 3.9 ( -8) Many Minority Managers & Supervisors [5-pt.J 3 . 1 3 . 6 3 . 6 2 . 3 3 . 6 2.4 Many Minority Managers & Supervisors [5-pt.J ( 1.2) ( 1.0) ( 1.0) ( 1.1) ( 1-0) ( 1.1) Availability of Stewards Who Speak First Language of Respondents [5-pt.] 3.4 4.1 3.9 2.3 4.0 2.3 Availability of Stewards Who Speak First Language of Respondents [5-pt.] ( 1.2) ( -7) ( 1-3) ( -9) ( -9) ( -9) Availability of Supervisors Who Have Same Ethnic Background [5-pt.] 3 . 0 3.3 3 . 4 2.4 3 . 4 2 . 4 Availability of Supervisors Who Have Same Ethnic Background [5-pt.] ( 1-2) ( 1.1) ( 1-4) ( i.o) ( 1.1) ( 1-0) Note. Rows of figures underlined are different at p_ < .05. Figures in brackets are sds. Table 5: Grievance Behaviours as a Function of Ethnic Grouping Group I Group II Group III Variable Overall Canadian European Canadian Visible Minority Majority Minority Job Satisfaction [15-pt.] 9.1 ( 2.0) 9.1 ( 2.3) 9.2 ( -9) 9.1 ( 2.0) 9.1 ( 2.1) 9.1 ( 2.0) Complaint to Supervisors [5-pt.] 2.8 ( 1-1) 2.9 ( 1.2) 2.7 ( -9) 2.6 ( 1.0) 2.9 ( 1-2) 2.6 ( 1-0) Complaint to Shop Stewards [5-pt.] 3.3 ( 1-0) 3.4 ( 1.1) 3.5 ( i.o) 3.2 ( 1-0) 3.4 ( l.D 3.2 ( 1.0) Complaint to Coworkers of 2.9 2.7 2.6 3.4 2.7 3.4 own ethnic group [5-pt.] ( 1.2) ( 1.2) ( 1-0) ( l.D ( 1.1) ( 1.1) Informal Settlement [10-pt.] 6.2 ( 1-6) 6.0 ( 1.6) 6.1 ( 1.3) 6.4 ( 1-8) 6.0 ( 1-5) 6.4 ( 1.8) Importance of Grievance Procedure [5 pt.] 4.3 ( -9) 4.4 ( -7) 4.5 ( -8) 4.1 ( l.D 4.5 ( .7) 4.1 ( 1-1) Fairness of Grievance Procedure [15-pt.] 9.3 ( 2.2) 9.0 ( 2.3) 9.8 ( 2.0) 9.7 ( 2.1) 9.1 ( 2.3) 9.7 ( 2.1) Instrumentality of Griev-ance Procedure [10-pt.] 6.1 ( 1.4) 5.8 ( 1.5) 6.5 ( 1-3) 6.3 ( 1.3) 5.9 ( 1-5) 6.3 ( 1.3) Producing Animosity After Grievance Filing [10-pt.] 6.3 ( 1.9) 6.3 ( 1-7) 6.2 ( 2.0) 6.5 ( 2.1) 6.2 ( 1-7) 6.5 ( 2.1) Worry of Management Reprisal for Grievance Filing [10-pt.] 6.4 ( 1.5) 6.6 ( 1.6) 6.5 ( 1-3) 6.1 ( 1.5) 6.6 ( 1.5) 6.1 ( 1.5) Note. Rows of figures underlined are different at p_ < .05. Figures in brackets are sd whereas those of the majority category are mainly English with a few European languages. Second, no minorities have ever been a shop steward. Third, minorities lived in other countries much longer than that of the majority. Most of minorities did not come to Canada until their teenage. Since they did not learn English as their first language, their self-reported English ability was lower than that of the majorities. Fourth, minorities reported seeing fewer minority managers, supervisors, and stewards (scored less than 3 on 5-point scales), whereas majorities reported seeing more of them (scored more than 3 on 5-point scales). Lastly, minorities also perceived a less positive multicultural atmosphere and discussed dissatisfaction more often with their ethnic coworkers. Categorizing respondents as minority and majority has reasons given the differences summarized in Tables 4 and 5. In particular, the two groups differ in years staying in other countries, English ability, and perceived demographic compositions in the workplace. Although categorizing respondents into three categories also seems meaningful, two considerations render further analysis unwarranted. One is that the European Canadian group has too few respondents (13), and the other is that some hypotheses of this study are related to differences between ethnic minorities and majorities. A focus on the minority - majority dichotomy fits the theoretical context of this study. Accordingly, subsequent analyses were focused on this dichotomy, whenever applicable. Analysis on the Three Components of the Multicultural Framework English ability, ethnic belonging, and atmosphere of multiculturalism were regressed on demographic variables, I-C values, demographic composition variables, and theoretically important interactions. The intercorrelations between all the dependent and independent 101 variables are shown in Table 6. English ability. Among the demographic variables, only the effect of schooling is significant (see column 1 of Table 7). Gender, age, and years in Canada are found to have no effects, although the latter two variables had been significant before other variables were entered into the equation. Among the socio-psychological variables, only the perceived proportion of minority coworkers in the workplace is significant. The regression coefficient indicates that the more minority workers one perceived in the workplace, the better the reported English ability. The reason may be that people regard themselves as better English speakers when more peers are minorities who probably speak only mundane English and, at times, speak other languages fluently in the workplace. The ethnic compositions of management and stewards have no effects. Together, the variables in the equation explain well the English ability differences between the minority and majority groups, as the group dummy variable is insignificant (p > .20). Ethnic belonging. People with high general collectivism report stronger belonging to their ethnic groups, although coworker I-C value has no effect. This finding supports the I-C Value Hypothesis (HI3) that group membership is more important to collectivists than to individualists. The group dummy variable is also significant and should be interpreted in light of the marginally significant interaction effect involving English ability. Ethnic status and its interaction with the group dummy are not significant, however. Understanding an interaction effect in regression analysis requires simple slope tests. These tests are done by deriving the simple regressions of, for example, ethnic belonging on English ability for the majority and minority groups (Aiken & West, 1991: 12). In other 102 Variable Table 6: Correlations between Variables 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 1. Complaint to Supervisor 2. Complaint to Shop 44 Stewards 3. Complaint to Coworkers of the same Ethnic group 19 21 -4. Job Satisfaction -35 -33 -38 -5. Informal Settlement by Shop Stewards 14 20 34 -11 -6. Ethnic Belonging -00 -08 04 01 18 -7. Atmosphere of Multiculturalism 18 02 -04 08 14 32 -8. English Ability 08 05 -14 01 -02 16 16 -9. Coworker Collectivism 19 07 24 -10 03 08 19 -06 -10. General Collectivism. 03 -12 -10 01 -09 23 25 18 19 -11. Ethnic Status -00 01 17 10 11 15 11 -09 20 -11 -12. Importance of Grievance Procedure 11 46 -02 -06 08 18 21 02 -03 06 10 -13. Fairness of Grievance Procedure 03 07 -04 39 03 08 23 -07 -09 08 28 20 -14. Instrumentality of Grievance Procedure 08 02 07 19 04 04 21 -08 12 -03 44 13 63 -15. Producing Animosity after Grievance Filing 03 06 26 -28 16 -00 -12 -12 11 14 06 -03 -08 -00 16. Worry of Reprisal from Management 07 18 13 -37 24 06 -02 00 05 08 -07 02 -36 -26 Note. Figures are Pearson correlation coefficients. Those underlined are significant at p_ < .01 (one-tailed test). Because of missing data in some items, N is between 108 to 90 (pairwise deletion). Table 6 (Continued): Correlations between Variables Variable 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 17. Proportion of Minority Managers -09 -02 -18 04 -02 11 13 10 -10 -04 -07 -07 -18 -11 -32 -06 -18. Proportion of Minority Coworkers 02 12 22 -13 09 04 10 08 04 15 24 19 12 10 06 -13 -16 -19. Proportion of Stewards Who Speak the First Language of Respondents 06 01 -11 02 05 12 19 35 -04 -04 -00 08 -11 -08 -12 16 37 03 -20. Proportion of Supervisors Who have the Same Ethnic Background -06 -03 -27 13 -07 17 30 14 -10 -01 -08 19 17 02 -06 04 31 -13 14 -21. Gender (Female = 1) -07 00 -05 15 -20 03 -08 03 04 09 01 -09 20 05 -00 -14 02 -10 -14 01 -22. Age -05 07 -03 06 -02 -02 -16 -10 -12 -02 12 16 14 02 05 07 -03 11 -06 02 -00 -23. Years of Schooling -15 -23 02 14 03 06 16 19 17 -00 -01 -23 -00 05 -02 -13 05 01 01 -01 -21 -16 -24. Years in Canada 02 08 -20 -00 -04 -04 02 13 -23 -12 -03 15 -05 -06 01 22 26 03 49 19 -12 55 -15 -25. Being As a Steward Last Year (Yes = 1) -31 -13 -06 18 -10 -03 -06 -13 -21 02 04 -04 15 05 -10 -18 -02 12 -14 00 20 15 -03 -03 Note. Figures are Pearson correlation coefficients. Those underlined are significant at p_ < .01 (one-tailed test). Because of missing data in some items, N is between 108 to 90 (pairwise deletion). Table 7: Regression Analysis on the Three Basic Multiculturalism Components Atmosphere of English Ethnic M u l t i -Variable A b i l i t y Belonging culturalism Gender . 13 ( ' 1. 11) .01( • 11) .08( .77) Age - . 18 ( -1. 14) .23 ( 1. 29) -.26 ( -1.73) Years of Schooling .26( ' 2. 31) .03 ( • 24) • 29( 2.67) Years i n Canada . 06 ( • 30) .41( -1. 78) .26( 1.31) As a Steward over Last Year . 16 ( '-1. 41) . 00 ( 01) -.06( - .61) Group (Majority/ Minority) -.27 '-1. 28) . 17 ( - . 70) .34 ( 1. 60) Coworker C o l l e c t i v i s m . 18 '-1. 56) .07 ( - . 56) .03 ( .32) General C o l l e c t i v i s m . 13 ' 1. 21) .22 ( 1. 81) .29 ( 2.74) Status of Ethnic Group . 04 '- . 33) . 02 ( • 13) . 30 ( 2.26) Many Minority Coworkers .29 ' i b 50) . 15. ( -1. 16) . 02 ( . 19) Many Minority Supervisors -. 09 [- . 76) . 01 ( • 04) .09 ( .72) Many Stewards who Speak F i r s t Language of Respondents . 19 [ 1-18) .20( 1. 35) . 15 ( .99) Many Supervisors who are of Own Ethnic Background . 12 91) .19( 1. I D .27 ( 2 .22) English A b i l i t y .23 ( 1. 33) .12 ( .83) Group x English A b i l i t y -.29( -1. 76) .09( .67) Group x Status .24 ( 1. 57) -.15( -1.13) Ethnic Belonging -.10 ( - .67) Group x Ethnic Belonging .48 ( 3 - 2 4 ) R2 .36 .29 • A l Adjusted R2 .23 . 11 .29 Note. The figures are standardized c o e f f i c i e n t s . Figures i n parentheses are t-values. Underlined p_< .10 Double Underlined p_< .05 (two-tailed tests) 105 words, it requires calculating the slopes of the regressions of ethnic belonging on English ability when the group dummy is equal to 0 (majority) and 1 (minority). The regression coefficients in the two simple regressions are then tested by t-tests with appropriate standard errors of the regression coefficients (see Endnote 1 of this chapter for detailed explanation). The simple slope analysis shows that ethnic belonging is positively related to English ability for the majority group, though only marginally significant, b ^ = .36, t(60) = 1.33, p_ < .10 (one-tailed test). In contrast, ethnic belonging is negatively related to English ability for the minority group, but is insignificant, b M i n = -.36, t(60) = -1.13. These findings suggest that better English ability is positively associated with ethnic belonging for the majority group, but is negatively, though insignificantly, associated with ethnic belonging for the minority group. Accordingly, the Linguistic Vitality Hypothesis (H4) is partially supported. The significant coefficient for years in Canada is caused by multicollinearity with age. Atmosphere of multiculturalism. Respondents perceived a more positive atmosphere of multiculturalism, the more schooling they had and the younger they were (see column 3). Females and males show no differences. These findings are consistent with studies both in Canada (e.g., Berry et al., 1977) and the U.S (e.g., Reitz & Breton, 1994: 30). Atmosphere of multiculturalism is also positively associated with general collectivism. To recall, the I-C value hypothesis (H14) suggests that collectivists, compared to individualists, regard harmony and consensus in the community (in this case the workplace) as more important and perceive the collective interest ahead of personal interests. Having these values may motivate them to value a multicultural atmosphere and see others as doing 106 the same. What is intriguing is the null effect of coworker I-C value. It seems that pursuing the collective benefits of coworkers over oneself, as governed by coworker collectivism value, is not related to perceiving multiculturalism in the workplace. Perhaps, it is important to have a recognition of collective interests involving other groups, such as the organization, the local management, and the union, rather than merely coworkers, for people to perceive multiculturalism in the practices and systems of the workplace. The general I-C value might have tapped the individuals' disposition of seeing beyond the interests of coworkers and their motivation of considering the interests of groups other than their peers in the workplace. Feelings of a high ethnic status is also associated with perceiving a positive atmosphere of multiculturalism. As suggested in the Ethnic Status Hypothesis (H12), having cultural security improves people's willingness to share and contact with other people (Berry, Kalin, & Taylor, 1977; Lambert & Taylor, 1990: 165). English ability exhibits no relationship with atmosphere of multiculturalism (H5), however. Ethnic belonging and the group dummy are not significant, but their interaction effect is significant. Simple slope analysis reveals that ethnic belonging is not related to the perception of an atmosphere of multiculturalism for the majority group, b ^ = -.08, t < 1. However, for the minority group, feeling stronger ethnic belonging is related to perceiving a positive atmosphere of multiculturalism, bM i n = .46, t(60) = 3.93, p < .01. These results, therefore, support partially the Ethnic Confidence Hypothesis (H7), which states that a strong ethnic belonging is related to the confidence of contacting and exchange with other groups. Lastly, seeing more supervisors of own ethnic background is associated with the perception of a positive atmosphere of multiculturalism. Other demographic composition 107 variables are not significant. This finding implies the possibility of having ethnically diverse supervisors, but not stewards or coworkers, in enhancing the perception of a multicultural atmosphere. Analysis on Grievance Behaviours Table 8 presents the results of the regression analysis on grievance behaviours. Demographic composition variables exerted no influence on any of the variables of grievance behaviours and were dropped from the final analyses. Only two demographic variables, years of schooling and last year as a steward, showed some effects on the criterion variables; other demographic variables were therefore dropped from the final equations as well. Job satisfaction. As shown in column 1, more schooling is related to higher job satisfaction. However, job satisfaction is negatively related to coworker collectivism and, marginally significantly, to the English ability of respondents. Hence, the I-C Value Hypothesis (H15) is supported. In addition, the opposite of the Linguistic Accommodation Hypothesis (H6a) are surprisingly supported. Speculatively, the reason may be that speaking better English equips workers, apart from sharing ideas, to disagree and argue with supervisors and management. The coefficients of ethnic belonging and atmosphere of multiculturalism are insignificant, but that of their interaction is significant. Follow-up analysis by simple slope tests suggests that ethnic belonging is positively associated with job satisfaction at one standard deviation above the mean of atmosphere of multiculturalism, b H = .30, t(60) = 1.75, p < .05. At the mean of atmosphere of multiculturalism (M = 7.2; ranging from 2 to 10), the relationship is still positive, but becomes insignificant, b M = .06, t < 1. At one 108 Table 8: Regression and Tobit Analysis on Grievance Behaviours Job Complaint Complaint Complaint No. of Grievances Variable Satisfaction (Supervisors) (Stewards) (Coworkers) Filed Years of Schooling .28( 2.55) -.28( -2.74) -.25( -2.31) -.01( - .12) - ,01( - .27) As a Steward over .11( 1.23) -.20( -1.92) -.02 ( - .18) . 04 ( . 38) -2.66 ( -4.14) Last Year .11( 1.23) -.20( -1.92) -.02 ( - .18) . 04 ( . 38) Majority/ Minority -.17( -1.48) . 03 ( .31) -.03 ( - .27) . 22 ( 1. 91) .37 ( 1. 29) Coworker Collectivism -.24 ( -2.16) .25( 2-35) .06( .51) . 30 ( 2. 56) . 11( 1. 64) General Collectivism -.02( - .77) .01( . 09) -. 18 ( -1.65) -.17( -1.51) - .14( -1.79) English Ability -.22 ( -1.80) • 21( 1.80) .13( 1.10) . 07 ( .58) . 22 ( 1-77) Ethnic Belonging .05( .45) -.13 ( -1.12) -.14( -1.17) .12( .99) . 10 ( 1. 19) Positive Atmosphere - .90) of Multiculturalism . 18 ( 1.45) . 09 ( .76) .11( .90) -. 02 ( - .17) - . 08 ( - .90) Belonging x -.01( - .09) -1.30) Multiculturalism .37( 2.99) -.21( -1.69) • 00( . 00) -.01( - .09) .06( -1.30) Job Satisfaction -.20( -1.88) -. 39 ( 3.52) -.28 ( -2.40) - .17( -2.43) Informal Settlement - .01( - .30) Complaint (supervisor) - .06( - .46) Complaint (Steward) . 12 ( .88) Complaint (Coworkers) - .23( -1.89) R2 Adjusted R2 . 12 .26 . 20 . 15 .48 Note. The figures are standardized coefficients. Figures in parentheses are t-values. Underlined p_< .10 Double Underlined p_< .05 (two-tailed tests) standard deviation below the mean, stronger ethnic belonging is associated with less job satisfaction, b L = -.19, t < 1, but is insignificant. Hence, the findings support the Secondary Multicultural Hypothesis (H9a). Complaint to supervisors. People who had more schooling tended to complain less frequently to their supervisors, and so did people who worked as a stewards over last year (see column 2). Satisfied respondents also complained less frequently to their supervisors. This finding is consistent with those of past studies (see Hla). However, coworker collectivism value and English ability are positively related to workers' frequency of complaints to their supervisors (H6b). Having better English skills helps the exchange between workers and supervisors. Moreover, as suggested by the I-C hypothesis (H17a) , individuals with high coworker collectivism are more inclined to confront their supervisors and managers than those with low coworker collectivism. Although ethnic belonging and atmosphere of multiculturalism exhibit no significant effects, their interaction does. Follow-up analysis by simple slope tests reveals that stronger ethnic belonging is associated with fewer complaints, b H = -.15, t(60) = -1.44, p_ < .10, at the one standard deviation above the mean of atmosphere of multiculturalism. At the mean of atmosphere of multiculturalism, the relationship still holds, but turns into insignificant, b M = -.07, t(60) = 1.12, n.s. At one standard deviation below the mean, the relationship drops to almost null, b L = .03. These findings again support the Secondary Multicultural Hypothesis (H9b). Complaint to shop stewards. As shown in column 3, people who had more schooling and more job satisfaction complained less frequently to their shop stewards (Hlb). 110 Complaining less frequently is also true for people with high general collectivism value. These people tend to value harmony and regard complaining to stewards as creating confrontations in the workplace (HI6b). Coworker I-C value shows no effects, and neither does English ability (H6c). Complaint to coworkers of own ethnic group. Discussing unhappiness with coworkers of own ethnic group is not related to education or being a steward over last year. However, job satisfaction is negatively related to the frequency of this discussion (Hlc). As predicted by the I-C Value Hypothesis, people with high coworker collectivism value the sharing with their peers in the workplace and more frequently talk about unhappiness with them (HI8). Moreover, minorities tend to talk about their unhappiness more frequently than majorities as the coefficient of the Group dummy variable is significant. As expected, English ability exerts no effect (H6d). Number of grievances filed. Tobit regression was used to study the number of grievances filed, for its distribution (M = 1-27; sd = 2.71; Max = 22) was truncated at zero, and OLS regression would give biased results (see Endnote 2 for further explanation). The mean grievance rate is high comparing to most workplaces. Higher grievance rates have been observed in grievance procedure studies because workers who have not filed a grievance tend not to respond to survey. They feel that they either lack the knowledge to answer the questions or find the grievance procedure unrelated to their works. Consequently, the mean grievance rates are usually upward biased. The potential effects of this bias will be discussed in Chapter 6. Readers should certainly take this into consideration when they interpret the results. i l l The results are shown in the last column of Table 8. Both job satisfaction and working as a steward over last years are negatively related to the number of grievances filed. The position probably required stewards to mind the "business" of other workers rather than their own. As expected, complaints to coworkers of own ethnic background is negatively related to the number of grievances filed (H2c). Talking about and sharing unhappiness in the workplace might reduce frustration and probably marshalled enough social support, so that estranged individuals did not file as many file grievances. Complaints to supervisors and stewards (H2a & H2b), however, display no significant effects. Better English ability is positively related to the number of grievances filed (H6e). The effect is beyond enhancing ethnic belonging and facilitating complaints, as these variables are in the equation. It may be that a better knowledge of English gives individuals a better idea about the formal grievance process. After all, people will not step into a risky decision if they do not know enough about the process they need to go through. As predicted, general I-C and coworker I-C values affect grievance filing in opposite directions. General collectivism value is negatively associated with the number of grievances filed (H16c), whereas coworker collectivism value is positively related to the number of grievance filed (H17c). These findings are consistent with the findings that collectivistic tendency can vary across target persons. To reiterate, "one can be very collectivist with regard to friends but totally independent and isolated from the family (Hui, 1988: 20)." In this case, the value of general collectivism disposes individuals to refrain from most confrontations and to seek harmony with other people; filing grievances in the workplace is 112 obviously against this value. However, the value of coworker collectivism motivates people to pursue the benefits of peers at the workplace because people strong in this value would see their coworkers as ingroups. As a consequence, they are willing to sacrifice themselves for their ingroups for collective benefits. Moreover, they would also regard people against them, such as managers and supervisors, as outgroups and, consequently, seek confrontation more eagerly than individualists may do under the same circumstance. The coefficients of ethnic belonging and atmosphere of multiculturalism are not significant, but their interaction is marginally significant. Simple slope tests reveal that at one standard deviation above the mean of atmosphere of multiculturalism, the slope is not significant, b H = .01, t(90) = .09. However, the slope becomes positive and significant at both the mean, b M = .32, t = 3.92, p < .01, and one standard deviation below the mean, b L = .63, t = 6.68, g < .01. These findings imply that ethnic belonging is positively related to the number of grievances filed when the atmosphere of multiculturalism is perceived as less positive. In other words, when individuals perceive a positive atmosphere of multiculturalism, a stronger ethnic belonging is not related to more grievance filing. In contrast, when the atmosphere is perceived not as positive, a stronger ethnic belonging is related to more filing. These findings provide further support for the Secondary Multiculturalism Hypothesis (H17e). Analysis on the Evaluations of the Grievance Procedure To recall, respondents evaluated the grievance procedure on five aspects: its importance, fairness, instrumentality to workers, feeling animosity after filing grievance, and worry of drawing reprisal from management. These variables were studied with respect to 113 the grievance behaviours, demographic composition variables, the basic components of multiculturalism, and I-C Value. Demographic variables displayed little effect and were dropped from the final equations. The regression results are presented in Table 9. Importance. The importance of grievance procedure is positively related to job satisfaction, complaint to supervisors, and complaint to stewards (see column 1). Also, people who used the grievance procedure over last year are likely to evaluate the system as important. Furthermore, two demographic composition variables are significant. People evaluated the grievance procedure as less important when they perceived the workforce as composing of more minority managers and supervisors. However, when they perceived more managers and supervisors who belonged to their ethnic groups in the workforce, they tended to regard the system as more important. Ethnic belonging and atmosphere of multiculturalism have no effects, but their interaction does. Simple slope analysis suggests that stronger ethnic belonging is associated with evaluating the procedure as less important at one standard deviation above the mean of atmosphere of multiculturalism, though insignificant, bH = -.07, t(71) = -1.05. At the mean, the relationship is null, b M = .02, but at one standard deviation below the mean, stronger ethnic belonging is related to evaluating the procedure as more important, b L = 2.65, t(71) = 2.65, p < .01. One possible interpretation is that when multiculturalism was not perceived as encouraged, people who felt strongly belonged to their ethnic group regarded grievance procedure as a means to protect themselves. If their opinions and behaviours might put them into trouble with other groups, a system that was more or less neutral might be useful to restore fairness when it was necessary. In other words, dismissing multiculturalism may 114 Table 9: Regression Analysis on Evaluations of Grievance Behaviours Variable Importance Fairness Instrumentality Majority/ Minority .07 ( .47) . 13 ( .79) .25| ' 2.11) Coworker Co l l e c t i v i s m . 11 ( -1 .14) -.10< '- .91) • 13 | [ 1.09) General C o l l e c t i v i s m . 12 ( 1 . 18) . 08 ( .74) -.18| [-1.45) English A b i l i t y • 111 '-1 . 13) -.10< ' - .88) -.03 ;- .23) Ethnic Belonging .05( .51) .10) .87) .04 [ -30) Po s i t i v e Atmosphere of Multiculturalism • 05< .46) . 16 ' 1 .40) .28 [ 2.27) Belonging x Multiculturalism .32 '-2 .90) -.03 ; _ .27) -.09 [- .67) Job S a t i s f a c t i o n .22 ' 2= ^ 1 ) .54 [ A . 66) .24 f 1.93) Informal Settlement . 04 r - .42) -.04 [- . 38) -.05 [- .43) Complaint (supervisor) . 18 ' l .71) .04 .34) . 13 [ 1.08) Complaint (Steward) .43 [ 3 .89) .31 [ 2 • 49) . 08 [ .59) Complaint (Coworkers) . 06 [- .57) . 13 [ 1 . 16) .03 [ .21) F i l i n g Grievance (Y/N) . 13 [ 1 .42) -.08 1 — .76) -.20 f-1.74) Many Minority Coworkers . 09 I 1 . 02) . 10 .96) Many Minority Supervisors .22 [-2 .14) -.13 [-1 . 09) Many Stewards who Speak my F i r s t Language . 08 .68) -.04 '- .32) Many Supervisors of Own Ethnic Background . 23 f 2 .22) . 18 [ 1 .60) R2 . 50 .39 .22 Adjusted R2 .39 .25 .08 Note. The figures are standardized c o e f f i c i e n t s . Figures i n parentheses are t-values. Underlined p_< .10 Double Underlined p_< .05 (two-tailed tests) 115 Table 9 (Continued): Regression Analysis on Evaluations of Grievance Behaviours Variable Animosity Reprisal Majority/ Minority -.35( -2 . 09) -.21( -1 .30) Coworker Co l l e c t i v i s m .03 ( .22) -. 07 ( - .67) General C o l l e c t i v i s m .24 ( . 12) . 18 ( 1 .71) English A b i l i t y -. 04 ( - .38) -. 03 ( - . 33) Ethnic Belonging -. 14 | -1 . 16) . 06 ( .49) Po s i t i v e Atmosphere of M ulticulturalism -.19 '-1 . 65) -. 09 ( .79) Belonging x Multiculturalism -.21 '-1 . 68) . 06 .54) Job S a t i s f a c t i o n -.31 '-2 . 65) -.50 '-4 .39) Informal Settlement . 17 [ 1 .54) .23 i 1 .21) Complaint (supervisor) -.14 [-1 . 15) -.14 [-1 .28) Complaint (Steward) -.12 [- .96) . 05 .40) Complaint (Coworkers) . 14 { 1 . 26) .00 . 02) F i l i n g Grievance (Y/N) -.07 . 68) -.13 [-1 .29) Many Minority Coworkers . 00 . 04) -.19 .88) Many Minority Supervisors -.44 (-3 .77) -.26 f-2 .33) Many Stewards who Speak my F i r s t Language -.12 ' — . 84) . 18 ( 1 . 31) Many Supervisors of Own Ethnic Background .09 .81) . 02 .18) R2 .36 .42 Adjusted R2 .22 .28 Note. The figures are standardized c o e f f i c i e n t s . Figures i n parentheses are t-values. Underlined p_< .10 Double Underlined p< .05 (two-tailed tests) 116 cause insecurity in workers, and leave the grievance procedure and a third party as their last resort. Fairness. Fairness is positively related to job satisfaction and complaint to supervisors, but not complaint to stewards (see column 2). No other variables are significant. Instrumentality. The overall model of instrumentality is not significant, probably because only a few predictors have only weak effects. A reduced model is marginally significant after the block of four demographic composition variables were removed (see column 3). This model, though post-hoc, implies that perceived ethnic composition in the workplace did not exist after the socio-psychological factors were taken into account. Introducing the demographic composition variables was redundant and cluttered the interpretation rather than helped. Despite the post-hoc nature, the analysis shows that job satisfaction and atmosphere of multiculturalism are positively related to perceived instrumental of the procedure. Minority workers tend to regard the procedure as more likely to produce instrumental outcomes than majority workers. However, filing a grievance is related to the belief that the procedure produces less instrumental results. I-C value has no effect on perceived instrumentality, although the signs are in the expected directions (H19a & H19b). Animosity. Feeling that filing grievance produces animosity between workers and supervisors is less likely if a person has higher job satisfaction, is a minority, and sees more minority managers and supervisors in the workplace (column 4). In contrast, it is more likely if a person talks about his/her unhappiness with his/her ethnic coworkers and has stronger general collectivism value (H20b). Coworker collectivism, however, exhibits no significant 117 effects (H20a). The interaction between ethnic belonging and atmosphere of multiculturalism is marginally significant. Follow-up analysis reveals that ethnic belonging is positively related to perceiving animosity at one standard deviation above the mean of atmosphere of multiculturalism, b H = -.28, t(71) = -1.62, g < .10. At the mean, the relationship still holds, but becomes insignificant, b M = -.14, t(71) = -1.16. The relationship approaches null, b L = .00, at one standard deviation below the mean. These findings suggest that the integrative function of an atmosphere of multiculturalism seems to be in effect. When the atmosphere is positive, feeling belonged to one's ethnic group reduces the feeling of animosity even though filing a grievance tests and threatens the relationship between workers and supervisors. In this sense, a positive atmosphere provides an integrative feeling or a buffer for people to deal with differences. This feeling, however, disappears as the atmosphere is perceived as less positive. Reprisal. Worry about reprisal from management is less likely among satisfied respondents (column 5). This worry is also less likely for workers who saw more minority managers and supervisors, and minority coworkers. However, having higher general collectivism value and seeing more informal settlement by stewards is positively related to worry of reprisal. People high in general collectivism are likely to believe that supervisors may express their anger by harassing them because, as shown above, they feel more animosity after filing grievances against the supervisors. The finding concerning informal settlement is unexpected, and will be discussed further in the next chapter. 118 Endnotes 1. A regression equation with an interaction term (the first equation below) can be restructured algebraically as shown in the second equation: Y= blA+b2B+ Y= b1A+b2B+ ... +bpX+bqZ+brXZ+b0 ... +(VV) X + (bqZ+bo> The coefficient of the regression of Y on X, (bp + br Z), depends on specific values of Z at which the slope is considered. Hence, (bp + b r Z) is dubbed the simple slope of the regression of Y on X at Z. Put it another way, a simple slope is the slope of the regression of Y on X conditional on a single value of Z. For instance, if Z is a dummy variable for the majority and minority grouping, simple slope at Z = 0 is the slope of Y on X for the majority. By the same token, simple slope at Z = 1 is the slope of Y on X for the minority. If Z is a continuous variable, such as ethnic belonging, there would be a family of simple slope regressions of Y on X , as Z can take on any values within its range. Meaningful levels of Z, such as median income level, can be studied by applying that particular value of Z to the equation. However, for psychological variables of which levels are more or less arbitrary, Cohen and Cohen (1983; cited in Aiken and West, 13) suggest using the values of Z M , Z H , and Z L , corresponding to the mean of Z, one sd of Z above the mean, and one sd below the mean, for studying the interaction. T-tests can be used to test the significance of simple slopes. The key is to derive the correct standard errors for (bp + br Z) when Z is given different values (Aiken & West, 1991: 24). It is known that the variances of any linear combinations of the original regression coefficients in an equation, U = w' b, is given by w'Sbw, where w is the vector of weights, and Sb the sample variance-covariance matrix of the regression coefficients. Using the above equation as example, the variance of the simple slope (bp + br Z) is equal to (s2,^  + 2 Z spr + Z 2 s2rr). s2pp is the variance of bp, s^ is the variance of br , and Sp r is the covariance of the estimates of bp and br. 2. Tobit model assumes that information for part of the dependent variable is missing, but the corresponding information for the independent variable is present. If OLS regression is used, the estimates would be biased. To estimate the true error term from the truncated distribution of the dependent variable, the maximum likelihood estimation is used with the assumption that the true dependent variable is normally distributed (Maddala, 1983). The Tobit procedure in Shazam (version 6.0) was used to analyze the data in this study. 120 Chapter 6: Discussion and Conclusion This study employed a modified social psychological framework of Canadian multiculturalism policy (Berry, 1984) to study the grievance procedure in an ethnically diverse organization. Hypotheses derived from this framework were tested by a survey on the blue-collar workers of a Crown corporation in the Vancouver plant. The findings confirmed major hypotheses derived from the framework. In the following I will first highlight the findings of the survey and, then, discuss their implications for research concerning individualism-collectivism, Canadian multiculturalism policy, grievance procedure research, and cultural diversity management. Lastly, I will qualify the findings and implications by discussing the limitations of the study. For the sake of illustration, findings will be interpreted as if the direction of causality is known although the correlational data of this study are unable to establish causality. In fact, significant results can always be interpreted from both directions. The discussion on findings about job satisfaction below illustrates these bi-directional interpretations as an example. Other results, however, will only be discussed in light of the causal directions implied in the multiculturalism framework. Summary of Findings Grievance behaviours. As expected, job dissatisfaction was found to be positively associated with complaint behaviours (Hla, Hlb, & Hlc) and the number of grievances filed (Hid). The apparent interpretation is that job dissatisfaction leads workers to complain more. Alternatively, as this study cannot confirm the direction of causality, another interpretation is that complaints lead to more job dissatisfaction. Complaining results in arguments with 121 supervisors or a recounting of unfair situations with stewards and coworkers, and, as a result, workers who make complaints may feel even more dissatisfied. These two interpretations are alternative explanations, and both could be reasons for the observed correlations. Complaining to coworkers of same ethnic group reduces the number of grievances filed (H2c). This implies that sharing frustration with and gaining social support from fellow workers of same ethnic background may reduce grievance filing. In contrast, complaining to supervisors and stewards do not seem to be the precursors of filing grievances as shown in the regression analysis (see Table 8). Yet, further analysis suggests another story. Dropping either one of the two complaint variables from the full model does not change the size of the coefficients. However, tobit regression that includes only one of the two complaint variables shows that the coefficients are positive and significant (p < .05). These findings suggest that the hypotheses (H2a & H2b) are correct, although the effects of complaints to supervisors and stewards are completely accounted for by other variables in the regression model shown in Table 8. Informal settlement by shop stewards was not related to the number of grievances filed (H3). Compared to studies of Bemmels and his colleagues (e.g., Bemmels et al., 1991), this study surveyed workers rather than shop stewards. Shop stewards may provide more accurate information on informal settlement than workers who only occasionally observe or hear about what stewards do. Alternatively, it may be that informal settlement does not work in this particular workplace due to the historical contest between the union and management. The latter possibility will be further discussed below. Linguistic vitality hypothesis. Results partially support the expectation that English 122 ability would enhance ethnic belonging. Specifically, white Canadians and Canadians of European descent feel a stronger belonging to their own ethnic group as their English ability improves. In contrast, better English ability does not, as hypothesized, reduce the ethnic belonging of visible minorities (H4). As most these individuals can speak and, at least, understand each other in their heritage language, learning a second language - English - does not seem to threaten their ethnic belonging as suggested in other studies (e.g., de Vries, 1990). Linguistic sharing hypothesis. English ability is expected to enhance intercultural sharing and, hence, the perception of a positive atmosphere of multiculturalism. However, results do not support this hypothesis (H5). The effect of English ability may not show up here because all of the workers had to be able to read and write English when they were hired. The variation in English ability is relatively little compared to immigrants in general. Linguistic accommodation hypothesis. There are mixed results for this hypothesis. As expected, better English ability increases complaints to supervisors (H6b) and the number of grievances filed (H6e). In addition, it also exhibits null effects on complaints to coworkers of same ethnic background (H6d). Against the expectations, however, it reduces job satisfaction (H6a), and is unrelated to complaints to shop stewards (H6c). Overall, these findings speak to when a mandated language is useful. English ability is important for talking to supervisors and managers and filling grievances, probably because the encounters entails good command of English for persuasion and argument, and the workers cannot rely on the other side to understand them. In contrast, English ability is not as important for conversing with coworkers of same ethnic background and, as found, shop stewards, because the workers can 123 speak their own language to their coworkers, and they can also rely on the others to understand their situations even if their English is not good. Ethnic confidence hypothesis. This hypothesis is partially supported (HI). In particular, ethnic belonging leads minority members to perceive a more positive atmosphere of multiculturalism in the workplace, but it does not lead majority members to perceive a negative atmosphere of multiculturalism. Therefore, a feeling of belonging to one's ethnic group, as found in Berry et al. (1977), provides security for minorities, whereas this belonging feeling, against the expectation of this study, does not result in ethnic complacency among majority members. Probably due to tolerance and multiculturalism beliefs ingrained in Canadian culture (Reitz & Breton, 1994), majority Canadians have little ethnocentrism and complacency, even if they feel strong belonging to their own ethnic groups. Primary and secondary multiculturalism hypotheses. The primary multiculturalism hypothesis receives no support from the data (H8a to H8e), although it was, unexpectedly, found to be enhancing for two aspects of the evaluations of the grievance procedure. Specifically, perceiving an atmosphere of multiculturalism leads people to evaluate the procedure as more instrumental to them, and to feel less animosity after filing grievances. Nonetheless, given the mostly null results, I consider it fair to say that perceived atmosphere of multiculturalism alone has only limited effects on grievance behaviours. Its weak effects are probably because grievance behaviours in a multicultural context are caused by the interaction of situational (atmosphere of multiculturalism) and dispositional factors (ethnic belonging). Findings concerning the secondary multiculturalism hypothesis support this possibility, 124 as the interaction of perceived atmosphere of multiculturalism and ethnic belonging affects job satisfaction, complaints to supervisors, and grievance filing in accordance with a priori expectations. When atmosphere of multiculturalism is perceived as high, a strong ethnic belonging enhances job satisfaction (H9a), and reduces complaints to supervisors (H9b) and, though insignificant, the number of grievances filed (H9e). When this atmosphere is perceived as low, however, a strong ethnic belonging leads to more grievance filing (HlOe), and, though insignificant, more complaints to supervisors (HlOb) and less job satisfaction (HlOa). In contrast, complaints to shop stewards (H9c & HlOc) and coworkers of own ethnic background (H9d & HlOd) are not affected by the interaction. Together, these findings provide support, though not overwhelmingly, to the integrative function of atmosphere of multiculturalism. Extrapolating from the literature of situational ethnicity (e.g., Okamura, 1981; Trimble, 1989), it is argued that individuals decide whether or not multiculturalism is encouraged in the workplace based on various environmental cues. If the answer is positive, the stronger their ethnic belonging, the more eager will they share their culture and communicate with other people. This translates to higher job satisfaction and fewer complaints. In contrast, if the answer is negative, the stronger their ethnic belonging, the more threatened they feel and the less they would communicate with others. This results in lower job satisfaction and more complaints. However, in circumstances where communication is not threatening, such as talking to coworkers and stewards, the atmosphere of multiculturalism and ethnic belonging do not seem to play a role. Ethnic status hypothesis. Feeling a high social status of one's ethnic group is found to 125 enhance the perception of an atmosphere of multiculturalism (HI2), but not ethnic belonging (Hl l ) . The former findings suggests that knowing that one's ethnic group is being respected makes a person perceive multiculturalism as being encouraged in the workplace. As argued, a higher ethnic status may prevent people from encountering discriminations against them, and, thus, give these people a pleasant state to help other people. Literature on altruism documents that people who are of high status and have good mood are more likely to supply help when the situations call for it (Schwartz, 1977). On the other hand, the null effect between ethnic belonging and ethnic status is unexpected. Perhaps, having an ethnic status was not perceived the same as winning a match or defeating an opponents in many of the studies concerning social identity theory (e.g., Brewer, 1979). Therefore, in the absence of this competitive backdrop, a higher ethnic status is not a strong factor in affecting one's ethnic identification. Individualism-Collectivism hypothesis. The results provide substantial evidence for this hypothesis, especially concerning the divergent effects caused by coworker and general I-C values. General collectivism increases ethnic belonging (HI3), atmosphere of multiculturalism (HI4), and perceived animosity after grievance filing (H20a), but it reduces complaints to shop stewards (HI6b) and the number of grievances filed (HI6c). On the other hand, coworker collectivism increases the numbers of grievances filed (H7c), complaints to supervisors (H7a), and complaints to coworkers of own ethnic background (HI8). To reiterate, two reasons explain the divergent and sometimes opposite relationships between grievance behaviours and the two I-C values. General I-C taps people's individualistic-collectivistic orientation across a number of target persons and social situations, 126 including friends, parents, and neighbours, whereas coworker I-C taps the orientation specifically towards coworkers. As a result, general I-C, but not coworker I-C, explains beliefs that are general and require a consideration beyond the workplace, such as ethnic belonging and atmosphere of multiculturalism. The second reason is that general I-C does not lead to acute ingroup-outgroup distinction in the workplace, whereas coworker I-C motivates workers to see such a distinction and, hence, regard supervisors and managers as outgroups. This sharp ingroup-outgroup distinction, in turn, motivates people to confront management, by filing grievances and complaining to supervisors, while perceiving less animosity concerning filing grievances. In contrast, general collectivism motivates people to avoid confrontation by talking their frustrations to their coworkers, filing fewer grievances, complaining less to shop stewards, and feeling more animosity after filing grievances. Evaluations of the grievance procedure. Job satisfaction is a consistent determinant for various evaluations of the procedure, though no formal hypotheses were made. Workers who are satisfied with the job evaluate the procedure as more important, fairer, more instrumental, less likely to cause animosity after grievance filing, and less likely to inflict worry of reprisal (Clark, 1988). Workers may generalize their satisfaction with the job to their evaluations of the grievance system. Moreover, complaining more to supervisors and shop stewards also leads workers to evaluate the procedure as more important and fair. One intriguing finding deserves further discussion. Worry of reprisal is positively associated with informal settlement of potential grievances by stewards, as observed by workers. An explanation is that worry of reprisal may lead workers to opt for informal settlement by steward, rather than filing grievances, and this desire results in the observation 127 of more informal settlement by stewards. Ethnic composition variables. Different from other studies on organizational demographics (Tsui & O'Reilly, 1989), this study employed perception variables rather than objective measures on group composition. Some interesting findings are found. Seeing more or fewer minority stewards in the workplace has no effects on any of the criterion variables. However, when more minority coworkers are seen in the workplace, a person tends to believe that he/she has better English ability and worries less about reprisal. Moreover, seeing more supervisors of own ethnic group leads to evaluating the grievance procedure as more important. The proportion of perceiving minority supervisors has more prevalent effects than the above variables as revealed in its coefficients in various regression equations. Seeing more minority supervisors may lead to perceiving a positive atmosphere of multiculturalism, evaluating the grievance procedure as less important, and worrying less about reprisal and animosity after grievance filing. It seems that increasing the proportion of minority supervisors would change several crucial perceptions of the workplace and the grievance procedure. However, as backlash on affirmative action programs suggests (Morrison & von Glinow, 1990; Taylor, 1991), simply increasing the numbers of minority managers is usually not the solution. Demographic variables. Years of schooling is an important demographic attribute. More schooling is related to better English ability, perceiving a more positive atmosphere of multiculturalism, feeling higher job satisfaction, and complaining less to supervisors and stewards. Conjecturally, schooling may give workers a different mindset towards work and 128 equip them with more skills to deal with difficult situations. Another finding is that older workers perceived the atmosphere of multiculturalism as less positive, which replicates the finding of Berry et al. (1977). Lastly, being a steward over the past year reduces people's complaints to supervisors and the number of grievances filed. The position may have led them to mind the problems of their members, and restricted them from complaining to supervisors about problems of their own. Ethnic differences. Majorities and minorities are different in three major aspects despite their similarities in age, working experiences, and other characteristics (see Tables 4 & 5). First, majorities spent much less time in a foreign country, whereas minorities usually arrived in Canada in their teens and, as a result, their reported English ability is poorer than the majorities. Second, majorities evaluate the workplace as more multicultural than minorities. Compared to minorities, majorities perceive a more positive multicultural atmosphere and a higher proportion of minority supervisors (Kossek & Zonia, 1993). The criteria used by both groups in judging ethnic diversity may be very different. Indeed, cultural myopia is used to describe dominant groups' low level of awareness of multicultural issues (Cox, 1993) and their inability to understand the institutional bias against women and ethnic minorities (Loden & Rosener, 1991: 52). See also Morrison and von Glinow (1990) and Kalin and Berry (1994). Third, other things being equal, minorities regard the grievance procedure as more instrumental and worry less about animosity after grievance filing than majorities. This finding may reflect the trust that Asians usually have towards authorities. Asians come from 129 high power distance cultures where superiors are respected, whereas Europeans and North Americans originate from low power distance cultures where superiors are distrusted and often challenged (Hofstede, 1991). Implications of Findings Individualism-collectivism. The basic tenet of I-C is held up in this study. The theory underlying I-C value postulates that people differ in the extent of integration with others and this individualistic-collectivistic tendency varies across and among persons (Triandis, 1989; Hui & Triandis, 1988). In social and cross-cultural psychology, some empirical works have already taken this point into account (e.g., Triandis et al., 1988). However, applied studies in business settings, except a few studies (e.g., Earley, 1993), seem to have ignored the crucial distinction between different kinds of I-C tendency, such as coworker and general I-C. This disregard may lead to the derivation of incorrect hypotheses and the use of inappropriate measurement of I-C values, such as, using general I-C scale to assess neighbour I-C value. To ameliorate the situation, researchers must consider what kinds of I-C values, such as parent I-C and coworker I-C, is important in their research and use appropriate measures to tap the kinds of I-C values supposed to be measured. Canadian multiculturalism policy. For most adults, work occupies about one-third of their time. The workplace is, therefore, a crucial venue where proponents of Canadian multiculturalism would like it to work. All tenets of the Canadian policy are sustained except the linguistic sharing hypothesis. However, it does not mean that the results of this study are entirely consistent with the present beliefs or findings of other researchers (e.g., Berry, 1977). 130 The chief distinction pertains to the interaction between ethnic belonging and the perception of an atmosphere of multiculturalism. The results indicate that maintaining ethnic belonging for people can lead to either negative or positive outcomes. What is crucial is whether people perceive a positive atmosphere of multiculturalism in the workplace. A positive atmosphere acts as either a protective cushion or an integrating force. It effects people who have strong ethnic belonging into positive work outcomes, such as higher job satisfaction and more security feeling after filing grievances. This person-situation interactional view is similar to findings based on the contact hypothesis. However, while Berry (1984) believes that maintaining ethnic identity alone would lead to other-group acceptance and integration, this study shows that it is the interaction between ethnic identity and the subjective social environment that makes the difference. If this interaction view is affirmed in future studies in other settings, the recommendations for multiculturalism policy makers would be to both (1) promote the maintenance of ethnic identity and (2) create a positive multicultural atmosphere by, for example, creating agencies of different ethnic origins to support minorities and immigrants. Therefore, although programs strive to maintain ethnic identity have recently been criticized as splitting the country (e.g., Spicer, 1991; Paquet, 1994), findings of this study, especially those related to the ethnic confidence hypothesis, demonstrate that promoting ethnic identification enhances perceived atmosphere of multiculturalism, which may, in turn, interacts with ethnic identification to fulfil a number of societal objectives. Promoting ethnic identity seems to be integrating people, and not splitting them. Grievance behaviours. This study demonstrates the importance of developing 131 alternative ways of solving conflicts in the workplace. People with low levels of coworker and high general I-C may dislike confrontation, which is the basis for typical grievance procedures. However, people high in general collectivism and low in coworker collectivism may prefer a procedure conducive to repairing damaged relationships rather than promoting confrontation. To them, dissipating animosity and restoring harmony is more important than knowing who is right in the dispute (e.g., LeResche, 1992; Leung & Wu, 1991). Emphasizing on restoring work relationships instead of determining rights is not new. Ury, Brett, and Goldberg (1988: 7) argued that a typical grievance procedure and arbitration are best in dealing with the issues concerning rights - social standards used by people to determine who is right and who is wrong. Rights are, however, rarely clear, and different standards, sometimes contradictory, can be applied to a particular situation. Reaching agreement on rights in the workplace, where the outcome will determine who gets what, can often be difficult, and frequently entails a third party and an arbitration. While some situations, such as ownership and redemption of reputation, call for the procedure to determine rights, more and more evidence (see below) has shown that the costs of arbitration and the use of grievance procedure to determine contractual rights in the workplace exceeds the benefits and usually the costs of using alternative resolution procedures. One source of costs is the time and attention needed in the pursuit of justice (Elliot & Goss, 1994: 58). Management and unions have to hire people to prepare for arbitrations and, to increase their chance of winning, may hire counsel to represent their interests. They also need to pay the arbitrator and witnesses in the longer and longer sessions of the arbitration. A survey by the Centre for Public Resources in New York surveyed companies that used 132 procedures other than arbitration between 1985-87. They found that sixty-one of the corporations and government agencies saved more than $49 millions in legal fees (News Release, 1988). The second source of costs is disputants' dissatisfaction towards the results of arbitration (Elliot & Goss, 1994: 59). Arbitration is seen as a judgment handed down by an "inquisitorial authority" based on what is heard about past conduct, actions, and interpretation. Questions concerning possible improvement and ways to improve are largely unaddressed. Consequently, the outcomes of arbitration may only touch the surface of a problem without getting at the root of it. The results are mutual dissatisfaction of the parties and recurrent conflicts in the workplace. Third, the nature of the arbitration process has turned business, economic, and management problems into legal issues (Elliot & Goss, 1994: 59). Workers and management whose cooperation will benefit each other cannot work out creative solutions because the concern of winning-losing forces the parties to inflate their demand, instead of working together to solve the problem. As a result, conflicts are debilitated, and cooperation is made even more difficult. Moreover, arbitrations require a long time to process. In Alberta, for example, an average case takes 11.5 months to be completed (Elliott & Goss, 1994: 10). Consequently, disputants do not find their grievance resolved and get on with their work, but have to worry about the outcome for a long time. In light of the above shortcomings, other procedures have been proposed to improve the efficiency of the grievance systems since the Alternative Dispute Resolution movement in the 60's (Goldberg, 1985). These procedures focus on preventing disputes, such as 133 compliance programs and quality of life programs, assisting arbitration, such as fact-finding and estimation of legal costs, and replacing arbitration, such as mediation and co-med (Elliot & Goss, 1994: 62). These procedures, in general, do not deal with rights, but reconcile the interests of the disputants. If they are used in the right situations, they can restore damaged work relationships and effectively solve problems in the workplace. Reconciliating interests is an approach which often costs less than determining rights in the work setting (Ury, Brett, & Goldberg, 1988). Interests are needs, desires, concerns, fears - things one cares about or wants. They underlie people's positions - tangibles or intangibles people say they want. Reconciling interests is not easy, however. It involves probing for deep-seated concerns, exploring creative solutions, and making concessions where interests are opposed. Mediation is widely used in labour relations to reconcile the interests of disputants (Goldberg, 1985). Mediation involves a neutral party facilitating the communications and negotiation of disputants, so that an settlement can be reached voluntarily. The mediator works either together with the disputing parties or privately with one of them in turn, and, though not often, may give them advice when necessary. The mediation process is flexible and allows the parties to structure the process themselves. Agreements, however, are made only by the parties, as the mediator does not adjudicate or impose a settlement on the parties. Studies show that mediation is superior to arbitration in reducing delay, cutting expenses, enhancing satisfaction of parties, improving work relationships and climate, and preventing the resurgence of similar cases (Elliot & Goss, 1994). In general, it has enjoyed high success rates in both the U.S. (Peterson, 1992) and Canada (Elliot & Goss, 1994, Chapter 5). 134 However, mediation and other alternative dispute resolution procedures enjoy much less success when dealing with conflicts in multicultural settings. Mediation services provided by many government and societal agencies are seldom used by minorities and, when used, are found to be dissatisfactory (LeResche, 1992; Duryea & Grundison, 1993). These procedures, such as mediation are set up based on the values of mainstream American or Canadian business practices, and they are conducted by trained mediators who are given a structure, suggested phrases, and ground rules that disputants are requested to follow. These procedures are flexible and very useful for a monocultural workplace, but they are rigid and almost useless for a multicultural workplace. Based on findings of LeResche (1992), for instance, Korean Americans may find it dissatisfying when a mediator, who conforms to the common practices, refuses to give them advice, and, instead, ask them to work out solutions with their disputants. The mediator may not know that he/she is in a position to advise the disputants according to the Confucius beliefs. Nonetheless, if the Korean Americans manage to continue the discussion, they may eventually find it dissatisfying when the so-far good discussion ends with the signing of an agreement without a recognition of mutual understanding and reestablishment of harmonious relationships through rituals, such as mutual apology and sharing a glass. Dissatisfaction arising from these issues can leave the Korean Americans puzzled about the success of the procedure and discouraged them from using the procedure again (see also discussion in Chapter 2). Therefore, existing mediation services and other alternative procedures, although generally successful, may need to be modified for accommodating the divergent ideas and 135 needs of different ethnic groups. Designing such a system is no easy task. One possible way is to follow the client-centred model proposed by Duryea and Grundison (1993). This model originates from their research on how five ethnic communities used and evaluated the government mediation services in Vancouver. They found that soliciting disputants in designing the resolution process is crucial for effective intervention. Accordingly, the intervention is not necessarily structured as involving the disputing parties sitting down at a table and following the ground rules spelled out by a mediator. Instead, intervenors learn in the process more about the perspectives of disputants and about the ways they can effectively address issues significant to them. Consequently, disputants and intervenors can structure the intervention process that suits the needs of the disputants. The model of Sheppard (1984) together with the cross-cultural literature on conflict-resolution (e.g., Ting-Toomey, 1988; Collier, 1991; Donohue, 1994) may form the basis for designing such a flexible, client-centred system in work settings, whereby disputants have more say on how the intervention process should be structured. An additional advantage of soliciting inputs from people of different ethnic backgrounds may be the enhancement of their perception of a multicultural workplace. According to the literature of procedural justice, giving participants a voice in structuring a system increases the perceived legitimacy and acceptance of the system (e.g., Sheppard et al., 1992; Leung, 1987; Leung & Lind, 1986; James, 1993). One reason for this acceptance is that having the chance to voice opinions strengthens people's feeling of being accepted by the group (Lind & Tyler, 1988; Tyler & Bies, 1990). This enhanced feeling of group acceptance is reminiscent of a positive multicultural atmosphere, wherein people believe they will receive 136 support from people of the organization regardless of their ethnic background. If this is the case, a flexible grievance system may benefit the workplace by strengthening the multicultural atmosphere in addition to more effective conflict-resolution. The null finding concerning informal settlement by stewards implies another way to improve the grievance system. Although this way is based on the findings of a particular workplace, it may well be applied to other organizations. It was speculated that informal settlement in this workplace is made impossible by the contentious union-management relationship. One steward gave the following answer when I asked her about the typical approaches of stewards, "...I could think of very few who had a problem-solving approach. Many of them just come to me and say that here is the collective agreement, and you violated this and that..." One steward did just that when I asked him to cite a case that came to his mind. Responding to my query, he flipped to a page of the collective agreement, pointed to an item about apron, and said, "...according to the agreement, a worker is entitled to have an apron...but when you asked for one from supervisors, they will argue against it..." According to him, the spirit of the contract was oftentimes not being carried out. From his point of view, it was the supervisors who did not try to solve the problems. He described the typical attitude of supervisors as, "...they know that the union does not have money and they try to drown you with money....they would say, ' if you don't like it, go ahead and grieve me.'" Since the interviewees were chosen by the union and management, the generality of their observations is unknown. In fact, some stewards may take a problem-solving approach to deal with so-called "soft" supervisors. Labelling himself as "soft," one supervisor noted, 137 "...having a good relationship with them [stewards] increases the tolerance and lower grievance filing...They would tell you what you did wrong instead of 'grieve' you." When probed about the ways to maintain a good relationship, he mentioned that "to treat them as individuals," "let them know what is going on," and "think and say things from their perspectives." Similar opinions were expressed by a long-term steward who recounted that "most supervisors are good [in their attitude]" when they approached her. She explained her good relationships with supervisors, "if you are reasonable with them and tell them where you are at, you will establish a reputation as reasonable." It seems that good and bad attitudes among stewards and supervisors reinforce each other once they happen. These excerpts suggest several changes that may facilitate more informal settlement by stewards. These changes, though based on data from this particular organization, may be applicable to other organizations that have a history of contentious labour relations. The first aspect is to train supervisors and stewards on how to handle conflicts with a problem-solving approach (e.g., Fisher & Reshef, 1992). At present, training on how to handle conflict and discuss with other parties is lacking. As one steward puts it, "We are on our own down there. No one told me how to handle a worker...It is emotionally draining and stewards have a very high burnout rate." As pointed out by many experts, handling conflict is never easy. Without adequate skills, minor conflicts may escalate to seemingly "irrational" proportions. While manuals are available for supervisors to learn to "say and do things at each step of the procedure," basic training on how to communicate and solve problems may prevent these manuals from being used and save valuable resources. Merely training the supervisors and stewards is by no means sufficient because, as put 138 by a shop steward, "supervisors are highly restricted by the management." Therefore, the fundamental approach of upper management has to be changed to collaborative and consultative as well (Cohen-Rosenthai & Burton, 1993). Despite its recent concern on labour relations (see Chapter 4), the focal organization is characterized by both stewards and supervisors as "authoritarian" and "traditionally acting like a dictatorship." There is the norm among management and supervisors, according to a steward, to be "threatening" and "being an asshole." Supervisors are "not supposed to teach people how to get things done but only to make sure that people do their things." Supervisors who attempt to establish a good relationship with workers are ridiculed by some others as "sucking up to the workers." Some management changes may alter the norms and help create a collaborative relationship. These measures may include joint union-management committees, profit sharing, and building quality-circle teams (Cohen-Rosenthal & Burton, 1993). Managing ethnic diversity. This study highlights two areas for diversity management. Stereotyping and misunderstanding about other ethnics is one aspect that needs to be addressed. One misunderstanding is about the fact that immigrants and minorities tend to cluster together "as little communities...they work together, dine together, socialize together and speak their own language." Oftentimes, majorities interpret these behaviours as creating fractions in the workplace. These stereotypic evaluations obscure the common knowledge that this is a way for some people to exchange social support and talk about dissatisfactions in the workplace. As a minority steward told me, "they do not complain so much to supervisors. They come to me and they complain a lot among themselves." In fact, while the clustering of minorities raises the eyebrows of the majorities and other ethnic minorities, 139 exchanging support with coworkers is more an attribute of collectivists, rather than an exclusive habit of minority workers. As shown in this study, it is not a function of ethnic membership but a function of I-C values. Another misunderstanding is that minorities and immigrants are submissive and follow orders all the time. When asked about their observations, one minority steward said, "it is true that you can push a Chinese further before he bounces back, but he bounces harder." Another steward expressed a similar view, "...they do not pursue things as often or have flimsy complaints. But when they have a complaint, they have a quiet tenacity...making sure that their wrong is righted." While these descriptions are generally true for Chinese, they are in fact the characteristics of collectivists. Collectivists try their best to preserve harmony with people, but would become eager to pursue things tenaciously once these people push too hard and cause collectivists classify them as outgroups. Hence, it is again a different way of dealing with things because of differences in I-C values. These misunderstandings highlights the need to inform people about I-C values, among other cultural issues, and behavioral differences between people who carry different values. This training may increase patience and understanding in interactions among people with different value orientations. It also helps people to become more tolerant and less ethnocentric. Bhawuk and Brislin (1991) succinctly discussed the characteristics of collectivists and individualists, and invented a scale to measure intercultural sensitivity concerning I-C. The scale is especially useful when it is used together with seminars and cultural simulators to improve intercultural communications (Brislin et al., 1986).1 'Cultural simulators are scenarios on intercultural interaction. 140 Perceiving an atmosphere of multiculturalism was found to exert positive influences on crucial organizational functions. Therefore, the second way to improve diversity management is to find ways to foster a positive atmosphere of multiculturalism. The findings suggest two ways. One is to select workers with high general collectivism, as this value is related to perceiving a positive atmosphere of multiculturalism. The second way is to increase the proportion of minority supervisors in the workplace, as the proportion is positively related to perceiving a positive atmosphere. However, the problem of causal direction discounts whether implementing these changes would exert positive effects. In addition, research on affirmative actions has shown that merely adding workers of different ethnic groups or genders may arouse backlash, and makes these new comers difficult to integrate with the original workforce (e.g., Taylor, 1991). A more systematic approach is, therefore, necessary, but beyond the scope of this study. As stated in Chapter 3, this study concerns only the symbolic, psychological aspects of multiculturalism, such as beliefs, attitudes, and feelings. It illustrates the socio-psychological mechanisms of multiculturalism in the workplace, and suggests the importance of promoting multiculturalism. However, programs and institutional changes that foster and drive the necessary psychological processes are beyond the purview of this study. Cox (1993) outlines a comprehensive approach to manage cultural diversity. He argues that atmosphere of multiculturalism in an organization should be regarded from three levels: structural, intergroup, and individual. There are different concerns at each level, and the ways to manage them are different. Structural characteristics include, for example, minority compositions in upper 141 management and the workshop level, recruitment procedures, and performance management. Intergroup level characteristics are, for example, mentoring programs and networking opportunities. Individual level characteristics are, for instance, stereotyping, ethnocentrism, and intercultural communications. As the three levels of characteristics are intertwined, a strong atmosphere of multiculturalism can only be achieved if managers take actions at all of the three levels in a concerted action. This study concerns mainly with the individual characteristics and partly with the intergroup characteristics, and it offers few implications on how to design a change program. Cox's model may be a more a useful platform for a successful plan of managing organizational diversity. Limitations This study is a one-time, cross-sectional survey. As noted above, this design does not prove the direction of causality. Statistically significant findings in the regression analysis merely establish these relationships. Obviously, more studies, especially longitudinal studies and quasi experiments, are needed to reexamine and to provide stronger evidence of causality for these findings. Two other concerns are related to the instruments of the survey. One issue is common method variance. Although the principal component analyses reported in Chapter 5 discounts this possibility, this problem cannot be completely ruled out. The interpretation of results of factor analysis is subjective. For one thing, there is no standards on what levels of variance explained by the first factor indicate serious common method variance (Podsakoff & Organ, 1986: 536). Future studies may take a more proactive approach and sidestep this problem by measuring constructs with different formats and assessing dependent and 142 independent variables at different times. Needless to say, these designs require fore-planning, funding, and substantial commitment of the research site. Another issue related to the instrument is about the validity of various scales, especially the I-C scales. The scales concerning grievance behaviours and evaluations of the grievance procedure seem to have expert validity, as they were approved by insiders of the research site. Likewise, the social belonging scale exhibits good psychometric properties similar to those in Phinney (1992). However, the validity of I-C scales is less certain. On one hand, the scales are derived from strong theory, and have probably the most vigorous validation among all available scales (e.g., Hui & Triandis, 1988). They were successfully applied in other studies as well. In fact, most of the a priori hypotheses pertaining to I-C in this study were supported, and hence the scales seemed to be measuring the latent constructs they are supposed to be measuring. On the other hand, the scales are generally low in internal consistency, and expectations concerning the divergent validity are not completely held up. One reason for these weaknesses is the relatively small numbers of items in the scales, which were kept to a minimal out of the concern of the length of the questionnaire. The other reason is that the I-C scale may be inappropriate for a sample composed of white Canadians, Asian immigrants, and middle-aged people. Although Hui and his colleagues (e.g. Hui, 1988) have validated the INDCOL across different cultures and samples, they used mostly students and local managers as their subjects and tested their scales in laboratory settings. The immigrant respondents and the work setting of this study might have affected the validity of the scales. Due to these uncertainties, readers should interpret findings concerning I-C values with caution. 143 The last limitation concerns the idiosyncracies of the research site and the generality of findings to other workplaces. As shown in other studies (e.g., Bemmels, 1994), grievance behavioral patterns vary significantly across industries as well as among different organizations and work groups within an industry. Hence, the findings of the study should not be extrapolated to other industries without careful considerations. Some characteristics of this site that deserve attention are as follows: (1) the site is a Crown corporation with a history of contentious labour relations; (2) sampled workers perform routine, boring tasks; (3) a new plant manager and the corporation as a whole were trying to improve the labour relations when the study took place; (4) the general economic conditions were stagnate, and they might affect the attitudes of the union and the management towards filing grievances; (5) the research site has a history of hiring minorities, especially Asian (estimated to be 25% to 50% of the workforce); and (6) because of the apathy of most Asian workers toward union activities, many of them do not participate actively in the union. Some of them, especially those who feel less satisfied with the union, might choose not to respond to the questionnaire and, consequently, bias the findings of the study. This potential bias deserves more discussion. The results should be generalizable to the whole sample, to the extent that the bias, if it exists, only shifts the means of various variables but not the relationships of various variables. In other words, the impact of the potential bias is a weaker observed relationships if it creates only a truncated range of observations. However, if the observed relationships based on this potentially biased sample are not the same across the whole sample, for example, the biased sample represents only the second half of a curvilinear relationship, the 144 generality of the results is highly restricted. It is hoped that further studies will be replicated in other plants of the Crown corporation, and other research strategies can triangulate the findings of this study (McGrath, Martin, & Kulka, 1982). Conclusion Bearing the above limitations in mind, this study enriches grievance procedure research with yet another theoretical framework - multiculturalism (see Figure 3) - in addition to exit-voice (Lewin & Peterson, 1988), behavioral models (Bemmels et al., 1991), attribution theory (Bemmels, 1991), and procedural justice (Klaas, 1989; Fryxell & Gordon, 1989). It also contributes to filling the cavity in the Canadian multiculturalism literature by supplying data concerning a multicultural workplace (Berry & Lapronce, 1994). An important message from the findings is that promoting multiculturalism in the workplace should pay off well in terms of raising job satisfaction, reducing the numbers of grievances filed, and enhancing the evaluations of the grievance procedure. Nonetheless, as ethnic diversity is primarily a new topic for labour relations research, it is hoped that this study would act as a springboard for further research. 145 References Aboud, F. E. (1981). Ethnic self-identity. In R. C. Gardner & R. Kalin (Eds.), A Canadian social psychology of ethnic relations (pp. 37-56). Toronto: Methuen. Adler, N . J., & Bartholomew, S. (1992). Academic and professional communities of discourse: General knowledge on transnational human resource management. Journal of International Business Studies. 23, 551-569. Aiken, L . S., & West, S. G. (1991). Multiple regression: Testing and interpreting interactions. Newbury Park, CA: Sage. Alderfer, C. P., & Thomas, D. A. (1988). The significance of race and ethnicity for understanding organizational behavior. In C. L. Cooper & I. T. Robertson (Eds.), International review of industrial and organizational psychology (Vol.3, pp. 1-41). New York: John Wiley & Sons. Allen, R. E. , & Keaveny, T. J. (1985). Factors differentiating grievants and non-grievants. Human Relations. 38, 519-534. Au, K. Y. (1993). Can satisfied employees curtail further commitment to a dubious investment? An exploratory study of Hirchman's (1970") argument. Working Paper, Faculty of Commerce, University of British Columbia. Beck, B. E. F., & Moore, L . F. (1985). Linking the host culture to organizational variables. In P. J. Frost., L . F. Moore, M . R. Louis, C. C. Lundberg, and J. Martin (Eds.), Organizational culture (pp. 335-354). Beverly Hills, CA: Sage. Bemmels, B. (in press). Exit-voice and employment relationships. In D. Lewin, D. J. B. Mitchell, & M . A. Zaidi (Eds.), Handbook of Human Resource Management. Greenwich, NJ: JAI Press. Bemmels, B. (1994). The determinants of grievance initiation. Industrial and Labor Relations Review. 47, 286-302. Bemmels, B. (1991). Attribution theory and discipline arbitration. Industrial and Labor Relations Review. 44, 548-62. Bemmels, B. (1988). Gender effects in discipline arbitration: Evidence from British Columbia. Academy of Management Journal. 31, 699-706. Bemmels, B. , Reshef, Y . , & Stratton-Devine, K. (1991). The roles of supervisors, employees, and stewards in grievance initiation. Industrial and Labor Relations Review. 45, 15-30. 146 Berry, J. W. (1980). Acculturation as varieties of adaptation. In A. Padilla (Ed.), Acculturation: Theory, models and some new findings (pp. 9-25). Boulder, CO: Westview. Berry, J. W. (1984). Multicultural policy in Canada: A social psychological analysis. Canadian Journal of Behavioral Science. 16, 353-370. Berry, J. W., Kalin, R., & Taylor, D. M . (1977). Multiculturalism and ethnic attitudes in Canada. Ottawa: Supply and Services Canada. Berry, J. W. (1991). Sociopsychological costs and benefits of multiculturalism. Working Paper No.24, Ottawa: Economic Council of Canada. Berry, J. W., & Laponce, J. A. (1994). Ethnicity and culture in Canada: The research landscape. Toronto: University of Toronto. Berry, J. W., Poortinga, Y. H . , Segall, M . H . , & Dasen P. R. (1992). Cross-cultural psychology: Research and application. New York: Cambridge University Press. Blank, R., & Slipp, S. (1994). Voices of diversity. New York: A M A C O M . Bochner, S., & Hesketh, B. (1994). Power distance, individualism/ collectivism, and job-related attitudes in a culturally diverse work group. Journal of Cross-Cultural Psychology. 25, 233-257. Bond, M . H. (1988). Finding universal dimensions of individual variation in multicultural studies of values: The Rokeach and Chinese Value Surveys. Journal of Personality and Social Psychology. 55, 1009-1015. Bond, M . H . , & Cheung, T. (1983). College students' spontaneous self-concept: The effect of culture among respondents in Hong Kong, Japan, and the United States. Journal of Cross-Cultural Psychology. 14, 153-171. Bond, M . H . , & Hwang, K.-K. (1986). The social psychology of Chinese People. In M . H. Bond (Ed.), The psychology of the Chinese people (pp. 213-266). Hong Kong: Oxford University. Borjas, G. J. (1991). Immigration research in the 1980's: A turbulent decade. In D. Lewin, O. S. Mitchell, & P. D. Sherer (Eds.), Research frontiers in industrial relations and human resources (pp. 417-46). Madison, WI: Industrial Relations Research Association. Borjas, G. J. (1994). Long-run convergence of ethnic skill differentials: The children and grandchildren of the Great Migration. Industrial and Labor Relations Review. 47, 553-147 573. Boroff, K. E. (1991). Measuring the perceptions of the effectiveness of a workplace complaint procedure. In D. Sockell, D. Levin, & D. B. Lipsky (Eds.), Advances in Industrial and Labor Relations (Vol.5, pp. 207-233). Greenwick, CT: JAI. Bourhis, R. (1983). Cross-cultural communication in Montreal: Two field studies since Bill 101. International Journal of the Sociology of Language. 46. 33-47. Boyacigiller, N . A . , & Adler, N . J. (1991). The parochial dinosaur: Organizational science in a global context. Academy of Management Review. 16, 262-290. Breton, R. (1978). Stratification and conflict between ethnolinguistic communities with different social structures. Canadian Review of Sociology and Anthropology. 15, 148-157. Breton, R., Isajiw, W., Kalbach, W., & Reitz, J. (1990). Ethnic identity and equality. Toronto: University of Toronto. Brewer, M . B., & Campbell, D. T. (1968). Ethnocentrism and intergroup attitudes. New York: John Wiley. Brewer, M . B. (1979). In-group bias in the minimal intergroup situation: A cognitive-motivational analysis. Psychological Bulletin. 86, 307-324. Brislin, R. W. (1986). The wording and translation of research instruments. In W. J. Lonner & J. W. Berry (Eds.), Field methods in cross-cultural research (pp. 137-164). Newbury Park, CA: Sage. Brislin, R. W., Cushner, K. , Cherrie, C , & Yong, M . (1986). Intercultural interactions: A practice guide. Newbury Park, CA: Sage. Burke, R. J. (1991). Managing an increasingly diverse workforce: Introduction. Canadian Journal of Administrative Sciences. 8, 62-63. Caddick, B. F. J. (1982). Perceived illegitimacy and intergroup relations. In H . Tajfel (Ed.), Social identity and intergroup relations (pp. 137-153). Cambridge: Cambridge University Press. Chaykowski, R. P., Slotsve, G. A. , & Butler, J. S. (1992). A simultaneous analysis of grievance activity and outcome decisions. Industrial and Labor Relations Review. 45, 724-737. Chan, D. K. S. (1991). Effects of concession pattern, relationship between negotiations, and 148 culture on negotiation. MA thesis. University of Illinois, Urbana. Christie, R., & Geis, F. (1970). Studies in Machiavellianism. New York: Academic Press. Cialdini, R. B. (1988). Influence: Science and practice. New York: Haper Collins. Cialdini, R. B. , Borden, R. J., Thorn, A. , Walker, M . R., Freeman, S., & Sloan, L. R. (1976). Basking in reflected glory: Three (football) field studies. Journal of Personality and Social Psychology. 34, 366-375. Clark, P. F., & Gallagher, D. G. (1988). The role of the steward in shaping union member attitudes toward the grievance procedure. Labor Studies Journal. 13, 3-17. Cohen-Rosenthal, E. , & Burton, C. E. (1993). Mutual gains: A guide to union-management cooperation. Ithaca, NY: ILR. Collier, M . J. (1991). Conflict competence within African, Mexican, and Anglo American friendships. In S. Ting-Toomey & F. Korzenny (Eds.), Cross-cultural interpersonal communication, (pp. 132-154). Newbury Park, CA: Sage. Connor, P. E. , Becker, B. W., Kakuyama, T., & Moore, L . F. (1993). A cross-national comparative study of managerial values: United States, Canada, and Japan. Advances in International Comparative Management. 8, 3-29. Cox, T. Jr. (1991). The multicultural organization. Academy of Management Executive. 5(2), 34-47. Cox, T. Jr. (1993). Cultural diversity in organizations: Theory, research and practice. San Francisco, CA: Berrett-Koehler Publishers. Cox, T. Jr., & Finley-Nickelson, J. (1991). Models of acculturation for intra-organizational cultural diversity. Canadian Journal of Administrative Science. 8(2), 90-100. Cox, T. H . , Lobel, S. A. , McLeod, P. L. (1991). Effects of ethnic group cultural differencs on cooperative and competitive behavior on a group task. Academy of Management Journal. 34, 827-847. Cox, T., Jr., & Nkomo, S. M . (1990). Invisible men and women: A status report on race as a variable in organizational behavior research. Journal of Organizational Behavior. 11, 419-431. Cummins, J. (1994). Heritage language learning and teaching. In J. W. Berry & Laponce, J. A. (Eds.), Ethnicity and Culture in Canada (pp. 435-456). Toronto: Toronto University. 149 Dalton, D. R., & Todor, W. D. (1982). Antecedents of grievance filing behavior: Attitude/behavior consisstency and the union steward. Academy of Management Journal. 25, 158-169. Defreitas, G. (1993). Unionization among racial and ethnic minorities. Industrial and Labor Relations Review. 46, 284-301. de Vries, J. (1990). Language and ethnicity: Canadian aspects. In P. L i (Ed.), Race and ethnic relations in Canada (pp. 231-250). Toronto: Oxford University Press. de Vries. J. (1977). Languages in contact: a review of Canadian research. In W. H . Coons, D. M . Taylor, & M . A. Tremblay (Eds.), The individual in language and society. Ottawa: Canada Council. Donnellon, A. , & Kolb, D. M . (1994). Constructive for whom? The fate of diversity disputes in organizations. Journal of Social Issues. 50, 139-155. Donohue, W. A. (1994). Communication issues in mediating cultural conflict. In J. Folger & T. S. Jones (Eds.), New directions in mediation: Communication research and perspectives. Thousand Oaks, CA: Sage. Dorroch, A. G. (1985). Another look at ethnicity, stratification, and social mobility in Canada. In R. M . Bienvenue & J. E. Goldstein (Eds.), Ethnicity and ethnic relations in Canada, (pp. 153-178). Toronto: Butterworth. Dreidge, L. (1989) The ethnic factor: Identity in diversity. Toronto: McGraw-Hill Ryerson. Duryea, M . L. (1993). Training intervenors to work cross-culturally. Cultural Diversity at Work. 6(2), 3. Duryea, M . L . , & Grundison, J. B. (1993). Conflict and culture: Research in five communities in Vancouver. British Columbia. Victoria, B.C.: UVic Institute for Dispute Resolution. Earley, P. C. (1993). East meets West meets Mideast: Further explorations of collectivistic and individualistic work groups. Administrative Science Ouaterly. 36: 319-348. Elliott, D. C , & Goss J. H . (1994). Grievance mediation: Why and how it works. Aurora, Ontario: Canadian Law Book. Elliott, J. L . , & Flearas, A. (1990). Immigration and the Canadian ethnic mosaic. In P. L i (Ed.), Race and ethnic relations in Canada (pp.51-76). Toronto: Oxford University Press. 150 England, G. W. (1975). The manager and his values: An international perspective from the United States. Japan. Korea. India, and Australia. Cambridge, MA: Ballinger. Erez, M . & Earley, P. C. (1993). Culture, self-identity, and management. New York: Oxford. Espinoza, J. A . , & Garza, R. T. (1985). Social group salience and interethnic cooperation. Journal of Experimental Social Psychology. 21, 380-392. Ewing, D. W. (1989). Justice on the job. Boston: Harvard Business School Press. Farrel, D. (1983). Exit, voice, loyalty, and neglect as responses to job dissatisfaction: A multidimensional scaling study. Academy of Management Journal. 26. 596-607. Fernandez, J. P. (1981). Racism and sexism in corporate life. Lexington, MA: Lexington. Fernandez, J. P. (1993). The diversity advantage. New York: Lexingtion Books. Fishbein, M . , & Ajzen, I. (1975). Belief, attitude, intention, and behavior: An introduction to theory and research. Reading, MA: Addison-Wesley. Fisher, E. G., & Reshef, Y. (1992). Shop steward performance: A constrained-decision analysis. Relations Industrielles. 47(1), 25-41. Foster, B. C , Jackson, G., Cross, W. E. , Jackson, B., & Hardiman, R. (1988). Workforce diversity and business. Training and Development Journal. April. 38-42. Freeman, R. B., & Medoff, J. L. (1984). What do unions do? New York: Basic Books. Fryxell, G. E. , & Gordon, M . E. (1989). Workplace justice and job satisfaction as predictors of satisfaction with union and management. Academy of Management Journal. 32, 851-866. Gandz, J., & Whitehead, J. D. (1989). Grievances and their resolution. In J. C. Anderson, M . Gunderson, & A. Ponak (Eds.), Union-management relations in Canada. Don Mills, Ontario: Addison-Wesley. Gans, H. J. (1979). Symbolic ethnicity: The future of ethnic groups and cultures. Ethnic and Racial Studies. 2, 1-20. Goldberg, S. (1985). The rise of grievance mediation. Proceedings. 37th Annual N . Y . U . Conference on Labor. New York: Mattew Bender. Goldstein, J. E. (1985). The prestige dimension of ethnic stratification. In R. M . Bienvenue 151 & J. E. Goldstein (Eds.), Ethnicity and ethnic relations in Canada (2nd ed., pp. 181-198). Toronto: Butterworths. Gordon, M . E. , & Fryxell, G. E. (1992). The role of interpersonal justice in organizational grievance systems. In R. Cropanzano (Ed.), Justice in the workplace (pp. 231-255). Hillsdale, NJ: Lawrence Erlbaum. Gordon, M . E. , & Miller, S. J. (1984). Grievances: A review of research and practice. Personnel Psychology. 37. 117-145. Gudykunst, W. B. (1988). Culture and intergroup processes. In M . H. Bond (Eds.), The cross-cultural challenge to social psychology (pp. 165-195). Newbury Park, CA: Sage. Haiven, L . (1994). Workplace discipline in international comparative perspective. In J. Belanger, P. K. Edwards, & L. Haiven (Eds.), Workplace industrial relations and the global challenge (pp. 70-102). Ithaca, NY: Cornell University. Hamilton, V. L . , & Sanders, S. (1992). Everyday Justice. New Haven: Yale University. Henderson, G. (1994). Cultural diversity in the workplace: Issues and strategies. Westport, CT: Quorum Books. Hinkle, S., Taylor, L . A. , Fox-Cardamone, D. L . , & Crook, K. F. (1989). Intragroup identification and intergroup differentiation: A multicomponent approach. British Journal of Social Psychology. 28, 305-317. Hofstede, G. (1980). Culture's consequences. Newsbury Park, CA: Sage. Hofstede, G. (1991). Cultures and organizations: Software of the mind. London: McGraw Hill . Hofstede, G., Bond, M . H . , & Luk, C.-L. (1993). Individual perceptions of organizational cultures: A methodological treatise on levels of analysis. Organization Studies. 14. 483-503. Hui, C. H. (1988). Measurement of individualism-collectivism. Journal of Research in Personality. 22, 17-36. Hui, C. H . , & Triandis, H. C. (1986). Individualism-collectivism: A study of cross-cultural researchers. Journal of Cross-Cultural Psychology. 17, 225-248. Hui, C. H . , & Yee, C. (1994). The shortened individualism-collectivism scale: Its relationships to demographic and work-related variables. Journal of Research in Personality. 28, 409-424. 152 Humphreys, L. G. (1949). Test homogeneity and its measurement. American Psychologist. 4, 245. Ibarra, H. (1993). Personal networks of women and minorities in management: A conceptual framework. Academy of Management Review. 18. 56-87. Isajiw, W. W. (1985). Definition of ethnicity. In R. M . Bienvenue & J. E. Goldstein (Eds.), Ethnicity and ethnic relations in Canada (2nd ed., pp. 5-17). Toronto: Butterworths. Isajiw, W. W. (1993). Definition and dimensions of ethnicity: A theoretical framework. Statistics Canada, & U. S. Bureau of the Census (Eds.), Challenges of measuring an ethnic world: Science, politics and reality (407-427). Washington, D. C : U.S. Government Printing Office. Jackson, S. E. (1992). Team composition in organizational settings: Issues in managing an increasingly diverse work force. In S. Worchel, W. Wood, & J. A. Simpson (Eds.), Group Process and productivity (pp. 138-173). Newbury Park, CA: Sage. Jackson, S. E. , & Associates (1991). Diversity in the Workplace. New York: The Guilford Press. James, K. (1992). The social context of organizational justice: Cultural, intergroup, and structural effects on justice behaviors and perceptions. In R. Cropanzano (Ed.), Justice in the workplace, (pp. 21-50). Hillsdale, NJ: Lawrence Erlbaum. Jamieson, D., & O'Mara, J. (1991). Managing workforce 2000. San Francisco, CA: Jossey-Bass. Johnston, W. B., & Packer, A. E. (1987). Workforce 2000: Work and Workers for the 21st century. Indianapolis, Ind.: The Hudson Institute. Kalin, R., & Berry, J. W. (1977). The social ecology of ethnic attitudes in Canada. Canadian Journal of Behavioural Science. 14, 97-109. Kalin, R., & Berry, J. W. (1994). Ethnic and multicultural attitudes. In J. W. Berry & J. A. Laponce (Eds.), Ethnicity and culture in Canada (pp. 293-321). Toronto: University of Toronto Press. Kanungo, R. N . (1982). Work alienation. New York: Praeger. Karambayya, R., & Brett, J. M . (1989). Managers handling disputes: Third-party roles and perceptions of fairness. Academy of Management Journal. 32. 687-704. Kemery, E. R., & Dunlap, W. P. (1986). Partialling factor scores does not control method 153 variance: A reply to Podsakoff and Todor. Journal of Management. 12, 525-30. Kim, K. I., Park, H.-J., & Suzuki, N . (1991). Reward allocations in the United States, Japan, and Korea: A comparison of individualistic and collectivistic cultures. Academy of Management Journal. 33, 188-198. Kirchmeyer, C , & McLellan, J. (1991). Capitalizing on ethnic diversity: An approach to managing the diverse workgroups of the 1990's. Canadian Journal of Administrative Science. 8, 72-79. Klaas, B. S. (1989). Determinants of grievance activity and the grievance system's impact on employee behavior: An integrative perspective. Academy of Management Review. 14, 445-458. Kluckhohn, C. (1951). Values and value orientations in the theory of action. Cambridge: Harvard University Press. Knight, T. R. (1986). Feedback and grievance resolution. Industrial and Labor Relations Review. 39, 585-598. Konrad, A. M . , & Gutek, B. A. (1987). Theory and research on group composition: Applications to the status of women and ethnic minorities. In S. Spacapan & S. Oskamp (Eds.), Interpersonal process (pp.85-121). Newbury Park, CA: Sage. Kossek, E. E. , & Zonia, S. C. (1993). Assessing diversity climate: A field study of reactions to employer efforts to promote diversity. Journal of Organizational Behavior. 14, 61-81. Kramer, R. M . (1991). Intergroup relations and organizational dilemmas: The role of categorization processes. Research in Organizational Behavior. 13, 191-228. Lambert, W. E. , & Taylor, D. M . (1990). Coping with cultural and racial diversity in urban America. New York: Praeger. Lautard, H . , & Clifton, R. A. (1990). The vertical mosaic revisited: Occupational differentials among Canadian ethnic groups. In P. L i (Ed.), Race and ethnic relations in Canada (pp. 189-208). Toronto: Oxford University Press. Lee, W. N . , & Tse, D. K. (1991). When cultures interact: Toward a socio-psychological model of consumption acculturation among ethnic minorities. Working Paper, University of British Columbia. LeResche, D. (1992). Comparison of the American mediation process with a Korean-American harmony restoration process. Mediation Quarterly. 9, 323-339. 154 Leung, K. (1987). Some determinants of reactions to procedural models for conflict resolution: A cross-national study. Journal of Personality and Social Psychology. 53. 898-908. Leung, K. (1988). Some determinants of conflict avoidance. Journal of Cross-Cultural Psychology. 19, 125-136. Leung, K . , & Bond, M . H. (1984). The impact of cultural collectivism on reward allocation. Journal of Personality and Social Psychology. 47, 794-804. Leung, K. , & Bond, M . H. (1989). On the empirical identification of dimensions for cross-cultural comparisons. Journal of Cross-Cultural Psychology. 20, 133-151. Leung, K. , & Lind, E. A. (1986). Procedural justice and culture: Effects of culture, gender, and investigator status on procedural preferences. Journal of Personality and Social Psychology. 50, 1134-1140. Leung, K. , & Wu, P.-G. (1991). Dispute resolution: A cross-cultural analysis. In R. W. Brislin (Ed.), Applied cross-cultural psychology (pp. 209-231). Newbury Park, CA: Sage. Lewin, D., & Peterson, R. B. (1988). The modern grievance procedure in the United States. New York: Quorum Books. Lewin, D. (1993). Conflict resolution management in contemporary work organizations: Theoretical perspective and empirical evidence. Research in the sociology of organizations. 12, 167-209. L i , P. S. (1990). Race and ethnicity. In P. S. L i (Ed.), Race and ethnic relations in Canada (pp. 3-17). Toronto: Oxford University Press. Lieberson, S. (1985). A societal theory of race and ethnic relations. In J. E. Goldstein & R. M . Bienvenue (Eds.), Ethnicity and ethnic relations in Canada (2nd ed., pp. 31-43). Toronto: Butterworths. Lieberson, S. (1993). The enumeration of ethnic and racial grops in the census: Some devilish principles. Statistics Canada, & U . S. Bureau of the Census (Eds.), Challenges of measuring an ethnic world: Science, politics and reality (pp. 23-35). Washington, D. C : U.S. Government Printing Office. Lincoln, J. R., & Miller, J. (1979). Work and friendship ties in organizations: A comparative analysis of relational networks. Adminstrative Science Quarterly. 24, 181-199. Lind, E. A. , & Tylor, T. R. (1988). The social psychology of procedural justice. New York: 155 Plenum. Loden, M . , & Rosener, J. B. (1991). Workforce America! Managing employee diversity as a vital resource. Homewood, IL: Business One Irwin. Lonner, W. L . , & Berry, J. W. (1986). Field methods in cross-cultural research. Newbury Park, CA: Sage. Maddala, G. S. (1983). Limited-dependent and qualitative variables in econometrics. Cambridge: Cambridge University. Major, M . M . , McCarrey, M . , Mercier, P., & Gasse, Y. (1994). Meanings of work and personal values of Canadian Anglophone and Francophone middle managers. Canadian Journal of Administrative Science. 11. 251-263. Mann, L . (1988). Cultural influence on group process. In M . H. Bond (Ed.) The cross-cultural challenge to social psychology (pp. 182-195). Beverly Hills, CA: Sage. Markus, H. R., & Kitayama, S. (1991). Culture and the self: Implications for cognition, emotion, and motivation. Psychological Review. 98, 224-253. Martin, J. E. , & Peterson, M . M . (1987). Two-tier wage structures: Implications for equity theory. Academy of Management Review. 30, 297-315. McCarrey, M . W., Gasse, Y . , & Moore, L. (1984). Work value goals and instrumentalities: A comparison of Canadian west-coast Anglophone and Quebec City Francophone managers. International Review of Applied Psychology. 33, 291-303. McGrath, J. E. , Martin, J., & Kulka, R. A. (1982). Judgment calls in research. Beverly Hills, CA: Sage. Messick, D. M . , & Mackie, D. M . (1989). Intergroup relations. In M . R. Rosenzweig & Porter, L . W. (Eds.), Annual Review of Psychology (Vol.40, pp. 45-81). Palo Alto, CA: Annual Reviews. Middleton, R. (1963). Alienation, race, and education. American Sociological Review. 28. 973-977. Mitchell, T. R. (1985). An evaluation of the validity of correlational research conducted in organizations. Academy of Management Review. 10, 192-205. Moore, L . F. (1995). Organizational responses to increasingly diverse workforces in British Columbia. Unpublished manuscript, University of British Columbia. 156 Morrison, A. M . , & Von Glinow, M . A. (1990). Women and minorities in management. American Psychologist. 45, 200-208. Multiculturalism and Citizenship Canada (1993). Directory of Canadian ethnocultural and bilateral business organizations. Ottawa: Supply and Services Canada. Nash, A. (1983). The union steward: duties, rights and status. Ithaca, NY: ILR. News Release (1988). CPR Legal Program, Survey of Legal Cost Savings. Nkomo, S. M . (1992). The emperor has no clothes: Rewriting "race in organizations." Academy of Management Review. 17, 487-513. Noble, B. P. (1993). Organizing the future. Harvard Business Review. July-August. 124. Oaker, G., & Brown, R. (1986). Intergroup relations in a hospital setting: A further test of social identity theory. Human Relations. 39, 767-778. Okamura, J. (1981). Situational ethnicity. Ethnic and Racial Studies. 4, 452-465. Omi, M . , & Winant, H. (1986). Racial formation in the United States: From the 1960's to the 1980's. New York: Routledge & Kegan Paul. Park, R. E. , & Burgess, E. W. (1921). Introduction to the science of society. Chicage: University of Chicago Press. Peterson, R. B. (1992). The union and nonunion grievance system. In D. Lewin, O. S. Mitchell, & P. D. Sherer (Eds.), Research frontiers in industrial relations and human resources (pp. 131-62). Madison, WI: Industrial Relations Research Association. Peterson, R. B. , & Fuller, E. (1992). China and Taiwan: Common culture but divergent economic success. Advances in International Comparative Management. 7, 185-201. Pfeffer, J. (1983). Organizational demography. In L. L. Cummings & B. M . Staw (Eds.), Research in Organizational Behavior (Vol. 5, pp. 299-357). Greenwich, CT: JAI. Phinney, J. S. (1990). Ethnic identity in adolescents and adults: Review of research. Psychological Bulletin. 108. 499-514. Phinney, J. S. (1992). The multigroup ethnic identity measure: A new scale for use with diverse groups. Journal of Adolescent Research. 7, 156-176. Pinder, C. C , & Stackman, R. W. (1995). Personal values and workplace relations: An update for practitioners and scholars. Vancouver: Centre for Labour-Management 157 Studies. Podsakoff, P. M . , & Organ, D. W. (1986). Self-reports in organizational research: Problems and prospects. Journal of Management. 12, 531-44. Porter, J. (1965). The vertical mosaic. Toronto: University of Toronto Press. Ray, R. (1994). Labour faces racism from within. Labour Times. 4(4), 1-2. Reitz, J. G., & Breton R. (1994). The illusion of difference: Realities of ethnicity in Canada and United States. Toronto: C D . Howe Institute. Roberts, L . W., & Clifton, R. A. (1990). Multiculturalism in Canada: A sociological perspective. In P. S. L i (Ed.), Race and ethnic relations in Canada (pp. 120-147). Toronto: Oxford University Press. Robinson, S. L . , & Bennett R. J. (1993). The four Ps of destruction: A multi-dimensional scaling study of deviance in the workplace. Working Paper #93:8, New York University. Rokeach, M . (1968). Beliefs, attitudes and values. San Francisco, CA: Jossey-Bass. Schneider, B. (1987). The people make the place. Personnel Psychology. 40. 437-453. Schwartz, S. H. (1977). Normative influences on altruism. In L. Berkowitz (Ed.), Advances in Experimental Social Psychology (Vol. 10). New York: Academic Press. Sheppard, B. H . (1984). Third party conflict intervention: A procedural framework. Research in Organizational Behavior. 6, 141-190. Sheppard, B. H . , Lewicki, R. J., & Minton, J. W. (1992). Organizational justice. New York: Lexington Books. Spicer, K. (1991). Citizen's forum on Canada's future: Report to the people and government of Canada. Ottawa: Supply and Services Canada, 27 June. Statistics Canada, & U. S. Bureau of the Census (1993). Challenges of measuring an ethnic world: Science, politics and reality. Washington, D. C : U.S. Government Printing Office. Stayman, D. M . , & Deshpande, R. (1989). Situational ethnicity and consumer behavior. Journal of Consumer Research. 16, 361-371. Syroit, J. E. M . M . (1991). Interpersonal and intergroup injustice. In R. Vermunt & H. 158 Steensma (Eds.), Social justice in human relations (Vol.1, pp. 259-277). New York: Plenum. Tajfel, H . & Turner, J. C. (1979). An integrative theory of intergroup conflict. In W. G. Austin & s. Worchel (Eds.), The social psychology of intergroup relations (pp. 33-47). Monterey, CA: Brooks/ Coles. Taylor, B. R. (1991). Affirmative action at work: Law, politics and ethics. Pittsburgh, PA: University of Pittsburgh Press. Ting-Toomey, S. (1988). Intercultural conflict styles: A face negotiation theory. In Y . Y. Kim & W. B. Gudykunst (Eds.), Theories in intercultural communication (pp. 213-235). Newbury Park, CA: Sage. Thiederman, S. (1991). Bridging cultural barriers for corporate success. Toronto: Lexington Books. Thomas, D. A. (1993). Racial dynamics in cross-race developmental relationships. Administrative Science Quarterly. 38, 169-194. Triandis, H . C. (1991). Theoretical concepts that are applicable to the analysis of ethnocentrism. In R. W. Brislin (Ed.), Applied cross-cultural psychology (pp. 34-55). Newbury Park, CA: Sage. Triandis, H . C. (1988). Collectivism vs individualism: A reconceptualization of a basic concept in cross-cultural social psychology. In C. Bagley & G. K. Verma (Eds.), Personality, cognition, and values (pp. 60-95). London: Macmillan. Triandis, H . C. (1989). The self and social behavior in differing cultural contexts. Psychological Review. 96, 506-520. Triandis, H . C , Kashima, Y . , Shimada, E. , & Villareal, M . (1986). Acculturation indices as a means of confirming cultural differences. International Journal of Psychology. 21, 43-70. Triandis, H . C , Leung, K. , Villareal, M . J., & Clack, F. L . (1985). Allocentric vs idiocentric tendencies: Convergent and discriminant validation. Journal of Research in Personality. 19, 395-415. Trickett, E. J., Watts, R. J., & Birman, D. (1994). Human diversity: Perspectives on people in context. San Francisco, CA: Jossey-Bass. Trimble, J. E. (1989). Multilinearity of acculturation: Person-situation interactions. In D. M . Keats, D. Munro, & L. Mann (Eds.), Heterogeneity in cross-cultural psychology (pp. 159 173-186). Amsterdam: Swets & Zeitlinger. Tsui, A. S., Egan, T. D., & O'Reilly, C. A. Ill (1992). Being different: Relational demography and organizational attachment. Administrative Science Quarterly. 37. 549-579. Tsui, A. S., & O'Reilly, C. A. Ill (1989). Beyond simple demographic effets: The importance of relational demography in superior-subordinate dyads. Academy of Management Journal. 32, 402-423. Tylor, T. R., & Bies, R. J. (1990). Beyond formal procedures: The interpersonal context of procedural justice. In J. S. Carroll (Eds.), Applied social psychology and organizational settings (pp. 77-98) Hilldale, NJ: Erlbaum. Ujimoto, K. V. (1991). Studies of ethnic identity and race relations. In P. L i (Ed.), Race and ethnic relations in Canada (pp. 209-230). Toronto: University of Toronto Press. Ury, W. L . , Brett, J. M . , & Goldberg, S. B. (1988). Getting disputes resolved. San Francisco, CA: Jossey-Bass. Vinokur, A . , Schul, Y . , & Caplan, R. D. (1987). Determinants of perceived social support: Interpersonal transactions, personal outlook, and transient affective states. Journal of Personality and Social Psychology. 53, 1137-1145. Wagner, J. A . , Ill & Moch M . K. (1986). Individualism-collectivism: Concept and measure. Group and Organization Studies, i i , 280-304. Watson, W. E. , & Kumar, K. (1992). Differences in decision making regarding risk taking: A comparison of culturally diverse and culturally homogeneous task groups. International Journal of Intercultural Relations. 16, 53-65. Watson, W. E. , Kumar, K. , & Michaelsen, L. K. (1993). Cultural diversity's impact on interaction process and performance: Comparing homogeneous and diverse task groups. Academy of Management Journal. 36, 590-602. Watts, R. J. (1994). Paradigms of diversity. In E. J. Trickett, R. J. Watts, & Birman, D. (Eds), Human diversity: Perspectives on people in context. San Francisco, CA: Jossey-Bass. Weinfield, M . (1994). Ethnic assimilation and the retention of ethnic cultures. In J. W. Berry & J. A. Laponce (Eds.), Ethnicity and culture in Canada (pp. 238-266). Toronto: University of Toronto Press. Wheeler, L . , Reis, H . T., & Bond, M . H. (1989). Collectivism-individualism in everyday 160 social life: The middle kingdom and the melting pot. Journal of Personality and Social Psychology. 57, 79-86. White, P. M . , Badets, J., & Renaud, V. (1993). Measuring ethnicity in Canadian censuses. Statistics Canada, & U . S. Bureau of the Census (Eds.), Challenges of measuring an ethnic world: Science, politics and reality (pp. 223-269). Washington, D. C : U.S. Government Printing Office. Zajonc, R. B. (1965). Social facilitation. Science. 149. 269-274. 161 Appendix A: Interview Guideline 1. Please describe an example of the most common type of grievance. 2. What are the issues that usually cause grievance? And why do they happen? Are there any ethnic differences? What kinds of grievances were filed because of these issues? 3. What are the relationships between workers and the management (the union) and between workers and supervisors (stewards)? Are there problems concerning motivational techniques, leadership style, discrimination, fairness and cooperation? 4. What procedures do you prefer to solve the grievance in the workplace? Do you think that people of different ethnic backgrounds may prefer different procedures? Why? 5. What options are available to people if dissatisfactions are not resolved through grievance procedures? 6. What are the possible consequences to workers and the company if grievances are not resolved or not resolved satifactorily? 162 Appendix B: Cover Letter of the Survey Package Appendix C: Questionnaire for Hypothesis-Testing Survey of Grievance Procedures: Beliefs And Use • This questionnaire asks for information about the grievance procedure, yourself, and people around you. • There are no right or wrong answers. We only need your honest responses. • Your participation is anonymous. The researchers, the union, and the management will not know if you return a completed questionnaire. • Please answer aU questions as best as you can. Section 1. Grievance Procedure A. How many grievances have you filed over last year? grievances B. If you have filed any grievances, how many of these grievances were about each of the following issues? Over Time Seniority , Use of Casuals Attendance Health & Safety Wage / Pay Discrimination Others(Specify) Please briefly describe the grievance (s): 1 2. Others C. When you answer the following questions, please circle the number corresponding to your answers. An example below shows how this is done. Strongly Strongly Disagree Disagree Neutral Agree Agree Example: My coworkers and I get along well 1 2 3 5 1. The performance of management is satisfactory 1 2 3 4 5 2. In general, I like working here J 2 3 4 5 3. Management do not treat me fairly 1 2 3 4 5 4. I often tell supervisors about my unhappiness of the workplace 1 2 3 4 5 5. I voice my work dissatisfactions to shop stewards 1 2 3 4 5 6. The grievance procedure is important for the workplace 1 2 3 4 5 166 Strongly Strongly Disagree Disagree Neutral Agree Agree 7. Union stewards discuss my complaints through informal discussion with my supervisor 1 2 3 4 5 8. I discuss my unhappiness at work with coworkers of my ethnic origin 1 2 3 4 5 9. Union stewards try settling my potential grievances with management before they file a grievance 1 2 3 4 5 10. The grievance procedure produces fair outcomes 1 2 3 4 5 11. The grievance procedure is designed in a fair way 1 2 3 4 5 12. Workers receive fair treatment from management while their grievances are being processed 1 2 3 4 5 13. Filing a grievance produces bad feelings among the disputing parties 1 2 3 4 5 14. Filing a grievance damages the relationship between management and workers 1 2 3 4 5 15. Management will try to retaliate if a worker files a grievance 1 2 3 4 5 16. Management does not harass workers after they filed a grievance 1 2 3 4 5 17. Those who file grievances usually get favourable outcomes 1 2 3 4 5 18. Workers usually gain good outcomes from grievance decisions 1 2 3 4 5 19. In my workplace, there are a large number of minority coworkers 1 2 3 4 5 20. In my workplace, there are only a few minority managers and supervisors... 1 2 3 4 5 21. People, regardless of their ethnic backgrounds, all support me in the workplace 1 2 3 4 5 22. Cultural diversity is not accepted in the workplace 1 2 3 4 5 23. Most of the union stewards who serve me at the workplace speak my first language 1 2 3 4 5 24. Few of my supervisors are of my ethnic background 1 2 3 4 5 25. One needs to return a favor if a coworker lends a helping hand 1 2 3 4 5 26. Assistance from coworkers is necessary for doing a good job 1 2 3 4 5 27. I would help a coworker who needs money to pay utility bills..... 1 2 3 4 5 28. If coworkers group together to help each other, they can gain a lot at no cost 1 2 3 4 5 29. I like to live close to my good friends -1 2 3 4 5 30. I know how to make friends with my neighbours 1 2 3 4 5 31. My ethnic group has a high status in society 1 2 3 4 5 167 Strongly Strongly Disagree Disagree Neutral Agree Agree 32. How people spend their income is not the business of their relatives (cousins, uncles) 1 2 3 4 5 33. I would not let my parents use my car (if I have one), whether they are good drivers or not 1 2 3 4 5 34. Even if a child wins a prestigious prize, the parents should not feel honoured in any way 1 2 3 4 5 35. People of other ethnic groups never discriminate against my own ethnic group 1 2 3 4 5 Section 2. Demographics I. Sex: M F 2. Age 3. Years of Education Yrs 4. length of service with present employer: Yrs 5. Years as a union member Yrs 6. Years in Canada: since birth Otherwise: Yrs 7. Contract Type: Part-time Full-time 8. Have you ever been a shop steward? Y N 9. I was a shop steward over last year? Y N 10. Job Nature: II. Please fill in: (a) My first language is ; (b) in terms of ethnic group, I consider myself to be Please circle the appropriate number for the following questions: strongly strongly Disagree Disagree Agree Agree 1. I am happy that I am a member of the ethnic group I belong to 1 2 3 4 2. I do not feel good about my cultural and ethnic background 1 2 3 4 3.1 have a strong sense of belonging to my own ethnic group 1 2 3 4 4. I speak well in English 1 2 3 4 5. I write well in English 1 2 3 4 Please write any comments, suggestions, or clarification for specific answers in the space below. Please make sure that you have answered all the questions, and return the questionnaire in the envelope provided. - Thank You -168 Appendix D: Questionnaire for Construct Validation Survey of Grievance Procedures: Beliefs And Use • This questionnaire asks for information about the grievance procedure, yourself, and people around you. • There are no right or wrong answers. We only need your honest responses. • Your participation is anonymous. The researchers, the union, and the management will not know if you return a completed questionnaire. • Please answer aU questions as best as you can. Section 1. Grievance Procedure A. How many grievances have you filed over last year? grievances B. If you have filed any grievances, how many of these grievances were about each of the following issues? Over Time Seniority Use of Casuals Attendance Health & Safety Wage / Pay Discrirnination Others( Specify) Please briefly describe the grievance (s): 1 2. Others C. When you answer the following questions, please circle the number corresponding to your answers. An example below shows how this is done. Strongly Strongly Disagree Disagree Neutral Agree Agree Example: My coworkers and 1 get along well 1 2 3 4 5 1. The performance of management is satifactory 1 2 3 4 5 2. The most important things that happen to me involve my present job 1 2 3 4 5 3. Most of my interests are centered around my job 1 2 3 4 5 4. I consider my job to be very central to my life 1 2 3 4 5 5. Many people around me provide me with encouragement 1 2 3 4 5 6. I receive useful information from a lot of people 1 2 3 4 5 170 7. Almost everyone says things that raise my self-confidence J 8. Many people listen to me when I need to talk.. 1 9. People around me show that they care about me as a person 1 10. My friends and my family understand the way I think and feel about things 1 11. Many union stewards in the workplace speak my first language 1 12. Many supervisors are of my ethnic background 1 13. One needs to return a favor if a coworker lends a helping hand 1 14. Assistance from coworkers is necessary for doing a good job 1 15. If coworkers group together to help each other, they can gain a lot at no cost 1 16. I would help a coworker who needs money to pay utility bills 1 17. I like to live close to my good friends 1 18. Management do not treat me fairly 1 19. Keeping my own cultural identity and characteristics is important 1 20. The best way to handle people is to tell them what they want to here 1 21. Anyone who completely trusts anyone else is asking for trouble 1 22. It is hard to get ahead without cutting corners here and there 1 23. It is important to share good and bad things with a friend even though s/he is not clever and usually cause trouble 24. The grievance procedure is important for the workplace 25. I know how to make friends with my neighbours 26. I like meeting and getting to know people from ethnic groups other than my own.... 27. I sometimes feel that it would be better if different ethnic groups didn't try to mix together. 28. I don't try to become friends with people from other ethnic groups... 29. I am active in organizations or social groups that include mostly memebers of my own ethnic group 30. In general, I like working here 31. I participate in cultural practices of my own ethnic group, such as special food, music or customs 171 Strongly Strongly Disagree Disagree Neutral Agree Agree 32. How people spend their income is not the business of their relatives (cousins, uncles) 1 2 3 4 5 33. I would not let my parents use my car (if I have one), whether they are good drivers or not 1 2 3 4 5 34. Even if a child wins a prestigious prize, the parents should not feel honoured in any way 1 2 3 4 5 35. People of other ethnic groups never discriminate against my own ethnic group 1 2 3 4 5 Section 2. Demographics I. Sex: M F 2. Age 3. Years of Education Yrs 4. length of service with present employer: Yrs 5. Years as a union member Yrs 6. Years in Canada: since birth Otherwise: Yrs 7. Contract Type: Part-time Full-time 8. Have you ever been a shop steward? Y N 9. I was a shop steward over last year? Y N 10. Job Nature: II. Please fill in: (a) My first language is ; (b) in terms of ethnic group, I consider myself to be Please circle the appropriate number for the following questions: strongly strongly Disagree Disagree Agree Agree 1. I am happy that I am a member of the ethnic group I belong to 1 2 3 4 2. I do not feel good about my cultural and ethnic background 1 2 3 4 3. I have a strong sense of belonging to my own ethnic group 1 2 3 4 4. I speak well in English 1 2 3 4 5. I write well in English 1 2 3 4 Please write any comments, suggestions, or clarification for specific answers in the space below. Please make sure that you have answered all the questions, and return the questionnaire in the envelope provided. - Thank You -172 Appendix E: Self-identified Ethnic Group, First Language, and Grouping S e l f - I d e n t i f i e d Group F i r s t Language Group Chinese Canadian (4)/ Canadian Chinese (2) White (9) Anglo-Saxon (2) / WASP (2) B r i t i s h / English (2) Chinese (14) Canadian (28) Hindustani (1) Hindi (1) Asian (2) Caucasian (11) I t a l i a n (1) Danish-Canadian (1) Indo-Canadian (1) Ukranian (1) F i l i p i n o (2) Canadian/ I t a l i a n (1) Ukranian/ Canadian (1) Canadian-Irish (1) / I r i s h Canadian (1) Scandinavian-Canadian (1) Hugarian (1) German (1) No-response (16) Chinese (5), English (1) III English (9) I English (4) I English (2) II Chinese (11), Cantonese (3) III English (25). Chinese (1) Janpanese (1), no-response (1) I English (1) III Hindi (1) III F i l i p i n e (1), Tagalog (1) III English (10), French (1 I English (1) II Danish (1) II Punjabi (1) III Ukranian (1) II Tagalog (2) III English (1) II English (1) II English (2) II Englsih (1) II English (1) II English (1) II I, III English (4) Chinese (1) F i l i p i n e (1). Cantonese (4), Mandarin ( l j , No-response (5)* Note. Figures i n brackets are the numbers of responses. Group I i s Canadian, Group II i s European Canadian, and Group III i s v i s i b l e minority. The f i r s t two groups are combined to form the Majority group, and group III are c a l l e d simply Minority i n the two-group system. *The 5 no-response cases are c l a s s i f i e d as Group III based on the years they stayed overseas. 173 

Cite

Citation Scheme:

        

Citations by CSL (citeproc-js)

Usage Statistics

Share

Embed

Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                        
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            src="{[{embed.src}]}"
                            data-item="{[{embed.item}]}"
                            data-collection="{[{embed.collection}]}"
                            data-metadata="{[{embed.showMetadata}]}"
                            data-width="{[{embed.width}]}"
                            async >
                            </script>
                            </div>
                        
                    
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:
http://iiif.library.ubc.ca/presentation/dsp.831.1-0087818/manifest

Comment

Related Items