This dissertation uses the method of historical narrative and the mechanism-process approach to study educational strategy selection. By presupposing that actions and decisions at a given point in time invariably affect subsequent actions and decisions, I adopted historical inquiry as the foundation of my study, as I believe it effectively explains how educational strategies are selected. To identify and analyze interactions and processes that shape educational strategies, I apply the case-study research design to specify objects, sites, subjects, and events involved. Finally, to provide a substantial body of empirical information to analyze, I draw on several data-collection techniques, including archival research, interviews, and the use of secondary analysis.
Research Question The research question central to this dissertation was born out of an inquiry that goes beyond observations of education opportunity, which are limited only to looking at the behavior and effects of governments on repressed groups. In order to broaden the analysis, I set out to answer a two-part question:
To what extent and how did the Jewish, Christian, and Baha’i communities in Iran select different strategies to meet educational needs under the Pahlavi and Islamic Republic regimes? To what extent and how does each group’s composition and characteristics, networks, and relation to both regimes affect their strategy selection?
Methodological Design In pursuing answers to these questions, this dissertation entails a chronological multicase study of three groups over two regime periods, covering 85 years of events and interactions. Furthermore, it is a study that uses an array of data sources and instruments, including archival sources, interviews, and secondary analysis. The methods of analyzing collected data are historical inquiry and interpretation through the mechanism-process approach. Finally, I use theoretical concepts from network analysis to examine existing network ties.
Sample Subjects and Sites Case subjects were selected from minority religious groups which are categorically similar, but which differ in their composition and characteristics, networks, and relations with regimes. Furthermore, religious communities have pre-existing organizational structures to oversee community affairs, making the task of following group strategies over a long period of time more manageable than strategies of temporary or event-based coalitions. I decided to include three cases because a comparative study would provide greater analytical leverage in explaining the influence of variations in group features bearing on educational strategy selection. Thus, I selected the Jewish, Christian, and Baha’i communities in Iran as diverse, but manageable, subjects.[1]
One of the central assumptions in this study is that regime-group interactions significantly shape group claims and actions. Thus, I choose to look at these three cases under two different regimes: the Pahlavi regime (1925–1979) and the Islamic Republic regime (1979–2009). Both regimes are characterized as high-capacity, authoritarian governments. However, each is distinguished by its ideological orientation, the former secular, the latter theocratic.
Sample Sources In this study three sampling strategies were employed depending on the source: (a) archival, (b) interviews, and (c) secondary analysis. I used mixed purposeful sampling for the three source categories.[2] Archival sources were the primary means of informing this study, including those available from governments, organizations, individuals, and the media. Furthermore, I conducted semi-structured and unstructured interviews with subjects living in Iran, Israel, the United States, Canada, and France. Finally, due to the scarcity of available and accessible information, secondary analysis sources became an important part of providing triangulation for other material and a means of providing information that was not accessed by the other two means.
Data Collection Archival Sources In dealing with a chronological case study spanning 85 years, archival information was the primary source of information for analysis. In the process of accessing governmental, organizational, individual, and media sources, and applying purposeful sampling, I created predetermined categories and subcategories for collection. Table A1[3] shows the general categorization and labeling scheme applied to sampling archival sources, which are described in greater detail below.
I sampled electronic sources in the above four categories by inputting key words into Internet search engines and word processing scanning applications. Because there were limited sources on the topic, I scanned for key words in every issue of those printed materials to which I had access. I used a hierarchical search technique, using primary key words, followed by secondary and tertiary key words input into search engines (details discussed below).
Purposeful and snowball sampling of printed sources was used by reading a wide range of secondary literature, identifying primary cited sources, and then seeking out those primary sources. Subsequently, I would search those primary sources for additional relevant information. Some sources were searched systematically and thoroughly, while others were accessed through general broad-based sampling using key word queries, employing an Internet search engine to locate additional sources. I was able to organize all material to provide sufficient information for triangulation and validation.
Archival instruments. In collecting the data, I used various technical instruments and skills, including translation, search engines, and research assistance. I used Persian and English sources for this study, drawing also on the work of a research assistant to expedite the reading of the Persian texts and identify relevant material. In looking for information sources and material, I developed a basic, flexible technique by applying hierarchical levels to key word sets in Internet search engines and word processing search features. I used a similar system for reading printed material.
Three general levels were used in search queries. Level one key words included case subjects and objects. Level two included primary themes of the study. Level three addressed specific issues, events, individuals and institutions, and other miscellaneous particulars arising from level two queries (varied and not systematic). These were then used in vertical and horizontal combination (e.g., Iranian Jews + education + Islamic Republic). Table A2 is a sample tabulation of the type of hierarchy of key word sets used to look for material directly related to this study. Tables A3 (Governmental Sources), A4 (Organizational and Individual Sources), and A5 (Media Sources) represent archival sources used in parts of this study. In addition to the four categories of archival sources mentioned earlier, the archival source tables specify the number of documents, types and location of sources, how they were accessed, and dates, when applicable.
Rationale. The sources selected for this study were chosen because, in addition to being useful in providing answers to the research questions, they were the most accessible to the researcher. They not only represent the kinds of sources that are generally used in studies of Iranian religious minorities, but in some cases include a wider range. Some sources which do not appear in the archival source tables in the Appendices, are found in the reference list, because they were not systematically retrieved, but rather resulted from unintentional referral or search. Other archival sources not included in these tables (or this study in general) are the result of limitations of the researcher (see section on limitations at the end of this chapter).
Interview Sources All three religious minority groups constitute vulnerable populations in Iran. It is not surprising that accessing information from members of any of the three groups was difficult, particularly because these groups have had to cope with restricted or risk-laden conditions by being circumspect in sharing information. Addressing the accessibility and availability of researching religious minorities in Iran, Eliz Sanasarian (2000), among the foremost experts in Iranian religious minority studies, asserts, “The scholarly literature on non-Muslim minorities is highly uneven, complex, and thinly researched” (p. 34). Knowing from preliminary research that access to human information sources would be challenging, and that information is sometimes distorted and generally guarded from out-group members, I set out to identify community leaders, organizers, and prominent members to be “key informants” for the study about community strategy selection. Key informants include individuals who were able to obtain “descriptive information that might be too difficult and time consuming to uncover through more structured data gathering techniques” (Blee & Taylor, 2002, p. 105). I initially employed purposeful stratified sampling by soliciting interviews from community religious leaders, organization representatives, and experts on each community, both inside and outside Iran. I defined the categories as follows:
1. Community religious leaders--rabbis (Jewish), clergy and pastors (Christian), Assembly members (i.e., NSA and LSA members, or equivalent for Baha’is)
2. Organization representatives--nongovernmental group organizations, community advocates participating in secular organizations, and committee members responsible for educational services and strategies
3. Experts--included academics and scholars who research one or more of Iran’s religious minority groups, nongovernmental and government agents who specialize in a particular group(s), lay members in the community who show significant knowledge about the group (i.e., journalists, rogue scholars, etc.)
I presumed that, without insider trust networks, I would not be able to effectively interview community leaders and members of groups inside Iran, and made contacts with individuals in Iran through contacts in the United States. Snowball sampling was also used in the interview process. In the course of my interviews with different group members (Jews, Christians, and Baha’is), I was referred to several members of the Iranian group communities who had experiences that were directly related to education in Iran. Through snowball sampling I was also able to access other community members who had experience with educational strategies used in Iran, including community run school administrators, teachers, students, and parents. Gender, age, and socioeconomic status were not included in sample design; only characteristics associated with roles in the community (i.e., community leaders, organization representatives, and prominent members) and educational experience in Iran was used in identifying subjects.
My initial target sample size was 10 community and organization leaders for each group, and three experts for each community; a total of 39 subject samples—given circumstantial restrictions I found this to be a sufficient number to provide essential information. With the goal of purposeful sampling of interview participants, I sent 19 independent requests for interviews with Jewish sources, 17 independent requests for interviews with Iranian Christian sources, and 14 requests for Baha’i sources (see Table A6). Contact information was obtained through Internet community websites, referrals by experts and other community leaders. These requests were sent by invitation letter (electronic) or in telephone calls. The acceptance rates were low. This was not surprising, given the high levels of tension in the social and political climate in Iran, and because of the sensitive nature of the topic and the understandable caution exercised by members of religious minority groups in general. In an attempt to offset the low response rates, I asked to be referred to others of similar rank; subsequently many pointed to the same source or directed me to an academic expert.
I was informed by several contacts that conducting research in Iran would not produce substantial results because of the “closed” environment. I was also warned that I should exercise caution because of the topic, as well as my own affiliation with an unrecognized religious minority group (i.e., Baha’i). However, after careful consideration, I decided— because of the dearth of existing information about Iranian religious minorities and educational issues —that it was necessary for me to travel to Iran to conduct anonymous, unstructured interviews. In Iran, I conducted interviews based solely on snowball sampling which began outside Iran through referrals, or while I was in Iran. I used extreme caution in interviewing members of Iranian religious minorities, not only because they are vulnerable, but also because of the sensitive nature of the dissertation topic.
Initially, I had set out to conduct five unstructured interviews with Iran-based informants involved in education of religious minorities for each group (i.e., students, teachers, administrators, etc.). However, my access to the minority school personnel I contacted—who will remain unidentified— was categorically denied. This clearly meant that approaching minority-run schools in the context of the current political climate in Iran was safe for neither the interviewee nor the interviewer. I was able to interview 28 Baha’is in Iran because of the insider sources I had already established prior to arrival in the country, but primarily because of the additional time that resulted when I was unable to access Jewish and Christian interviewees. I was able to interview two Christians and three Jews living in Iran. Most of those whom I interviewed wished to remain anonymous. Tables A6 and A7 reflect the sampling numbers by purposeful sampling and snowball sampling (based on an extensive interview database that was created to keep track of interviews by group, organization affiliation, position/role, location, type of communication, and dates).
Interview instruments. An interview guide was prepared and approved by the Institutional Review Board (IRB) at Teachers College, Columbia University, to be used in interviews with community and organization leaders, prominent community members, and topic experts (Table A8 shows the interview guide). The guide was nominally adjusted depending on the individual’s role and group affiliation, and translated into Persian. The interview guide included questions that solicited descriptive and explanatory answers related to their group and educational strategies. The guide included questions related to each group’s educational opportunities and challenges in Iran (past and current), as well as the group’s characteristics and composition, networks, and regime relations bearing on educational opportunities and strategies. I remained flexible and adjusted the interview guide according to each participant’s receptivity and expertise with the topic. I also left time for the participants to share information they felt would inform the study. This was particularly the case with those who did not know much about educational issues in the community, but were knowledgeable about other matters bearing on education, such as regime-group relations and community characteristics and composition.
I also employed unstructured interviews for follow-up session with interviewees, as well as all Iran-based participants. The usefulness of unstructured interviews was apparent when participants shared information about issues which did not appear in any other primary or secondary source, or about which I had no previous knowledge. I incorporated several themes in unstructured interviews, particularly questions related to motives, fears, considerations in strategy selection, as well as experiences of opportunity and challenges related to educational pursuits. In addition, at some point during the unstructured part of the interview, I asked about other individuals whom I could contact. I also included written communication as part of the unstructured interview process, usually in the form of two or three questions soliciting answers about a particular subject, source, or fact. These were equally important, especially when retrieved from a key informant.
Interviews were conducted in person and using communication devices, in both English and Persian. In most interviews outside Iran, I used a voice recorder and took written notes. In Iran, I took only encrypted notes, and would then write out and email expanded notes to myself (still using cryptic language). I kept a ongoing database, into which all interviews were logged by participant descriptor when available (i.e., gender, age, initials to remind me who it was), date, position/occupation, organization/notes, date of interview, modality (i.e., written correspondence, personal communication, or telecommunication), and location.
Interviewees reside in various cities around the world, including Canada, France, Iran, Israel, and the United States. I spent 2008–2010 in Los Angeles, California as my main location of field research, because most members of Iranian religious minorities living outside Iran live in Southern California—particularly Jews, Armenian-Iranian Christians, and Baha’is. I traveled to Iran in December 2009 to conduct interviews. The political climate in Iran during this period was unstable as a result of the post-election protests and riots of 2009.[4] However, I was able to travel and conduct interviews in Tehran, Shiraz, Babol, and Sari during my stay.
Consent and Confidentiality Given the vulnerability of group members and the sensitive nature of the study topic, I extended anonymity rights to all those whom I interviewed (unstructured and semi-structured). Most interviewees chose to remain anonymous. I followed IRB regulations for semi-structured interviews. I obtained oral consent and verbally conveyed stipulations of confidentiality with interview participants to reduce potential risk to participants in the study. The consent form included the purpose of the study, the researcher’s background (personal and academic information, and contact information), and confidentiality particulars. All participants were specifically informed that they could stop the interview at anytime, and did not have to answer any questions about which they felt uncomfortable.
Secondary Analysis and Sources Secondary analysis and literature was an important part of this study, as it provided important background information on settings, regimes, and subjects. The information regarding religious minorities in Iran is scattered, fragmented, and sometimes unreliable (Sanasarian, 2000). One contribution which this dissertation makes to the study of religious minority groups lies in the efforts to synthesize and critically analyze the existing primary and secondary information in this field. I used several academic secondary analyses, cited throughout, to complete a comprehensive narrative of otherwise fragmented streams and episodes describing educational experiences of religious minority groups. The use of primary sources and secondary analysis also reinforces triangulation of validity of data (see validity issues in section below). I attempted to access as many secondary analyses of Iranian religious minority community issues related to education. Much of the secondary source material gathered for all three groups repeats information and often cites other similar sources. This was useful in identifying the limits and strengths of secondary literature.
Data Analysis Methods Theoretical Propositions I used preconfigured theoretical propositions to categorize collected data.[5] The study follows a historical timeline with distinct actors in question. Thus, I arranged the information along these two dimensions. The two major categories/themes are regimes and religious minorities, in order to optimally organize the data for a chronological multicase study, using the mechanism-process approach. Three other preconfigured codes were also assigned to religious minorities, intrinsically related to the research question: (a) group composition and characteristics; (b) group networks; and (c) group-regime relations. Finally, the category education was used to identify information directly related to educational opportunities, challenges, and strategies in modern Iran. Figure 4 illustrates the basic organization of data by category. The categories and labels were used to organize and manage collected data from archival sources, interviews, and secondary analysis.
Figure 4. Preconfigured proposition for categorizing data.
Interviews Analysis After collecting recordings or notes from interviews, I scanned the content and used the preconfigured thematic codes (see Figure 4) to label interview content. After listening to the recorded interviews, information was included which was either missing or incomplete in my corresponding interviews notes. When information was shared that was outside the bounds of the categories, I would measure its perceived importance by how it would contribute to the study and labelled it “miscellaneous (short description).” Not all content derived from interviews was of equal length or quality. However, because of this labeling scheme (dividing up interviews by religious minority group), I was able to access interview content easily when looking at other data for the given group and time period.
Four distinct categories emerged in the data collected in interviews:
1. Historical--events (episodes, interactions, etc.)
2. Descriptive--community characteristics, statistics
3. Relational--relations with the government and other religious minorities; networks within and outside the community
4. Referential--additional potential people to interview, and additional sources
Most interviews were used in the process of constructing historical events for each period, but also to inform particulars about how characteristics, networks, and regime relations affected educational and other social strategies. In many cases information was supplemented by archival and secondary sources, but also used to supplement archival sources that were incomplete. Interviews added a rich texture to the study, particular for the Islamic Republic period, during which so little has been written about Iranian religious minorities.
It would be appropriate to mention here that I used secondary interviews in my study, but treated them as archival sources. I used a similar process of labeling those interviews using my theoretical propositions. Of particular value were the Center for Iranian Jewish Oral History and the Foundation for Iranian Studies Oral History Program. Short interviews available in documentary films contributed additional empirical evidence for the study.
Historical Analysis Historical inquiry. Using preconfigured codes listed in Figure 4, I set out to find a substantial body of primary sources to outline the historical development of educational processes which could then be broken down into bound streams and episodes (see mechanism-process approach below). I employed historical inquiry and interpretation as my initial means of analyzing and organizing this information. Historical inquiry is not just telling a story. Rather, it is the critical analysis of sources from various periods in the past, interpreting events and interactions, arranging them in logical sequence, with the goal of explaining what, how, and why a phenomenon in question occurred in the past. The categorized material was then systematically organized so as to be readily accessible when I moved through a chronological sequencing of events relating to educational development in Iran, for the religious minority groups specifically. Part of the process of analysis was to determine the validity and quality of the collected sources. In most cases, I was able to triangulate the data with other primary or secondary sources. However, in others cases, the information gathered was all that was available to me, and I had to use discretion in relying on those sources. In general, depending on the type of source, I used various preanalytical methods in drawing out useful information to create a historical narrative, against which I could apply the mechanism-process approach.
Document and media analysis. Concepts from document analysis were identified by themes in the texts. Those (particularly media) which contained the religious minority titles (i.e., coded as any derivation of Jew, Christian, Armenian, Assyrian, Chaldean, and Baha’i) were classified using a filing system. For print media (Ettelaat, Iran Times and Kayhan), I examined all issues for these primary codes. For raw archive material, such as letters, reports, and pictures, I collected information using the same classification scheme as in Figure 4, noted the content that related to the study, and filed it under the appropriate category. I was able to effectively examine how group composition and characteristics, networks, and regime-group relations affected strategy selection by using historical inquiry and interpretation of events and interactions within this designated framework. I did not engage in a quantitative analysis of the documents nor do an in-depth discourse analysis of the text. The purpose of the document analysis was to construct a holistic and balanced historical narrative which could be further analyzed through the mechanism-process approach. It should be noted that in Chapter 5, I discuss the particular features of group characteristics and composition, networks, and regime-group relations, including definitions of concepts.
Mechanism-Process Analysis This study set out to explain how a group meets its educational needs. While I discussed the theoretical background of the mechanism-process approach at length in the literature review, it is important to consider how it applies directly to this dissertation. I employed the mechanism-process approach in Chapter 6 in order to identify specific streams and episodes of interaction that significantly influenced educational strategies and their outcome. In the process of historical analysis using various sources, I specifically looked for streams and episodes of contention and actuation. As Tilly and Tarrow (2007) describe, “episodes are bounded sequences of continuous interaction, usually produced by an investigator’s chopping up longer streams of contention into segments for purposes of systematic observation, comparison, and explanation” (p. 36). I also use the terms streams of actuation and episodes of actuation to denote regime-group interactions characterized by facilitation and tolerance. Although more unwieldy than contentious interactions, streams and episodes of actuation highlight important periods of strategy development for the case subjects when conditions and government education policies are more favorable than contentious. Based on available data for each period (i.e., Pahlavi and Islamic Republic), streams and episodes related to education opportunities and challenges become the units of analysis in studying educational strategy selection. I identified episodes of contention and actuation for the three religious minority groups. Often, several smaller events or micro-episodes were identified and analyzed within episodes. Once episodes were identified and described, I proceeded to apply the eight steps in the mechanism-process approach as outlined by Tilly and Tarrow (2007, p. 207, summarized in Table A9).
The purpose of analyzing my historical interpretation through the mechanism-process approach was to explain a) how strategies were actually selected, developed, and deployed, and b) how such strategies affected the subsequent selection of other strategies in meeting educational needs. In Chapter 2 (Literature Review), I defined essential terms such as mechanisms, processes, and episodes. Due to variation in the extent of information available and accessible for each time period and across the different groups, including general differences in group features, multiple scales of observation were used to identify processes. To identify mechanisms and processes over time, I borrowed the categories developed by Tilly and Tarrow (2007), and interpreted strategies in those terms (see Table A10 for a listing and definitions of mechanisms and processes used to explain strategy selection during analysis in Chapter 6). I also include specially created content-specific processes when analyzing the episodes, and describe these in greater detail in Chapter 6. While processes were identified and reassembled to explain the selection and development of strategies, I also interjected an interpretation of how each group’s composition and characteristics, networks, and regime-group relations influenced the shaping of those processes.
Concepts of Network Analysis In Chapter 5, concerning the religious minority groups’ composition and characteristics, networks, and regime-group relations, I used concepts (not mechanics) of network analysis to visualize the domestic and international network ties of each group. Instead of using specifically designed surveys or computation of archival data—input into a network analysis software program, as typically used for network analyses—I made a simplified and holistic estimate of network ties and their values, based on a range of qualitative and quantitative sources available to this author.
Five nodes were classified for inclusion in the network mapping. For each group, I looked at following prominent actors:
1. Local community organizations and members
2. Transnational community organizations and members
3. Nongovernmental organizations (nongroup)
4. Governments and their agencies
5. International and supranational organizations (nongovernmental and governmental institutions)
I initially identified node ties with the subject group by labeling archival sources, interviews, and secondary literature that mentioned such connections with a code for network tie (i.e., NT). Nodes or actors were identified by geographical location. Since technical network analysis methods are beyond the scope of this dissertation, I only use approximated measures to determine the values of network ties. The sources I use to determine network nodes, tie strength, tie relation, and tie type, include government and organization documents and sources, organizational source material, news media sources, in-depth interviews with organizational leaders (15 participants inside and outside Iran), and other archival and secondary sources (see Tables A2, A3, and A4 for the list of sources used). Four features are included in my conceptual network mapping exercise: network nodes (i.e., actors); types of ties between nodes; relational content of ties; and strength of ties between nodes.
I narrowed the focus on three other basic features as part of my conceptual network analysis:
1. Types of relational ties between nodes—direct, indirect, independent
2. Relational content of ties—information, resources, advocacy
3. Strength of ties—between nodes
Three types of relational ties are identified in my approximation. Direct ties indicate direct exchange or interaction between an Iranian group community and another node (whether in or outside Iran). An indirect tie indicates that one of the three interactions (informal, advocacy, or resources) exists, but through an additional party (i.e., broker) working as a conduit of exchange. The term independent ties indicates existing network ties between different entities (particularly governments), representing state-state relations or organization-organization cooperation. Drawing on international relations theory, I argue that these relations have an indirect impact on minority groups who are associated with communities in those particular countries.
Examined here are those ties involving the exchange (unidirectional or bidirectional) of information, advocacy, or resources.[6] Information refers to that which can be utilized for reports and assessments, not information in the sense of resources used toward community development (e.g., curriculum or school models). I use the term resources to refer to the categories of Edwards and McCarthy (2004; see literature review), including material, moral, socio-organizational, human, and cultural. Advocacy refers to any form of initiative that denotes explicit protest or support of particular regime policies and practices, or group actions; some forms of advocacy include mutual agreements, petitions, letters of appeal, voiced concern, motions made in multilateral supranational organizations, and sanctions. In the network maps in Chapter 5, icons at the end of the connecting lines identify content ties (information, resources, and advocacy).
To estimate the measure of tie strength, I use the descriptive values of weak (s = 1), moderate (s = 2), strong (s = 3). It is especially important to emphasize that network strength was estimated by accounting for recorded frequency of interaction, level of actions (i.e., prominence), and longevity of the relationship. Strength is not computed scientifically, but is attributed to emerging trends that surface from review of primary and secondary sources. In the network maps in Chapter 5, strength is represented by the thickness of the lines connecting nodes.
Validity Issues Several research methodologists have suggested that to reduce threats to the validity of a given study, triangulation of data collection ought to be incorporated into the study (Creswell, 2007; Maxwell, 1996; Yin, 2009). In this study I used archival sources of various types (media journals, letters and memoirs, organization documents and publications, government documents and publications), semistructured and unstructured interviews, and secondary analysis to provide adequate triangulation.
Nevertheless, some consideration of potential threats to validity is warranted due to general research bias, interview participant bias, and source bias. My background and affiliation with an Iranian religious minority group, could be considered a potential threat to the study. Growing up in a community that framed the situation facing the Baha’is in grave terms created a preconception of the situation facing religious minorities. This was offset in a number of ways. By adopting the idea that all regimes engage in repression, toleration, and facilitation, I moved out of the fallacious dichotomous approach to government behavior (i.e., repressive or nonrepressive). More importantly, in using Baha’i sources or interviews, I exercised extreme discretion by triangulating data from non-Baha’i sources (i.e., governmental and nongovernmental, and more specifically from academic expert analyses). Conducting interviews with vulnerable religious minorities also poses some challenges. Validity of information gathered from these interview participants could have also been compromised when participants were fearful of having information used against them or others. Here again, I relied on triangulation to confirm information, using other interview sources, archival sources, and secondary analysis. Additionally, key informants were extremely helpful in compensating for the low number of interviews. Finally, there are sources on Iranian religious minorities (whether primary or secondary), and some are clearly biased. Thus, lack of information may also have given rise to nuances that were not accounted for in this study. However, I am confident that through the use of multiple sources, major threats to validity were overcome.
Limitations The limitations to this study can be categorized as those related to language, access, safety, and current climate in Iran. I used Persian and English to conduct the study, but was unable to use sources in Hebrew, French, Armenian, and Syriac. As mentioned earlier, conducting research on religious minorities in Iran is an extremely difficult task because of the sensitivity of the situation and the reticence of these groups since the Islamic Revolution. The caution and insularity of Iranian religious minority communities has only been amplified by current political and social restrictions during the presidency of President Ahmadinejad (2004–2009), the increase in religious minority harassment and slander in media, and the post-election protests and government crackdown since 2009. Thus, compromised trust and restricted conditions by the government inhibited freer access to more sources. The need for constant vigilance and concerns for safety were other limitations. I did not want to put my interviewees or myself at risk, particularly those participants who live or have connections in Iran. Similarly, I was advised by group informants that Christians and Jews are extremely insular in Iran, and even those who have recently left are reluctant to share information with outsiders. Notwithstanding these limitations, by using a wide range and differing types of sources, I was able to collect sufficient information to undertake this essentially unstudied research.
[1] I did not include the Zoroastrian community because of small numbers, limited access to sources, and time constraints; i.e., there are only some 10,000 Zoroastrians in Iran, living in mid-size cities and rural areas (Sanasarian, 2000).
[2] According to Johnson and Christenson (2008), “mixed purposeful sampling is likely to be used when a researcher uses data triangulation—examining multiple data sources which might be selected according to different sampling methods” (p. 246).
[3] See Appendix A for all Tables in this chapter.
[4] Two days before my arrival there had been mass protests in Tehran, and on the day of my departure there was an escalation of violence in Tehran.
[5] With regard to case studies, Yin (2009) writes: “The first and most preferred strategy is to follow the theoretical propositions that led to your case study. The original objectives and design of the case study presumably were based on such propositions, which in turn reflected a set of research questions, reviews of the literature, and new hypotheses or propositions. The propositions would have shaped your data collection plan and therefore would have given priorities to the relevant analytical strategies” (p. 130).
[6] There may be other dynamic interactions present, but for the sake of brevity and this heuristic exercise, I have retained focus on only these three features.
Research Question The research question central to this dissertation was born out of an inquiry that goes beyond observations of education opportunity, which are limited only to looking at the behavior and effects of governments on repressed groups. In order to broaden the analysis, I set out to answer a two-part question:
To what extent and how did the Jewish, Christian, and Baha’i communities in Iran select different strategies to meet educational needs under the Pahlavi and Islamic Republic regimes? To what extent and how does each group’s composition and characteristics, networks, and relation to both regimes affect their strategy selection?
Methodological Design In pursuing answers to these questions, this dissertation entails a chronological multicase study of three groups over two regime periods, covering 85 years of events and interactions. Furthermore, it is a study that uses an array of data sources and instruments, including archival sources, interviews, and secondary analysis. The methods of analyzing collected data are historical inquiry and interpretation through the mechanism-process approach. Finally, I use theoretical concepts from network analysis to examine existing network ties.
Sample Subjects and Sites Case subjects were selected from minority religious groups which are categorically similar, but which differ in their composition and characteristics, networks, and relations with regimes. Furthermore, religious communities have pre-existing organizational structures to oversee community affairs, making the task of following group strategies over a long period of time more manageable than strategies of temporary or event-based coalitions. I decided to include three cases because a comparative study would provide greater analytical leverage in explaining the influence of variations in group features bearing on educational strategy selection. Thus, I selected the Jewish, Christian, and Baha’i communities in Iran as diverse, but manageable, subjects.[1]
One of the central assumptions in this study is that regime-group interactions significantly shape group claims and actions. Thus, I choose to look at these three cases under two different regimes: the Pahlavi regime (1925–1979) and the Islamic Republic regime (1979–2009). Both regimes are characterized as high-capacity, authoritarian governments. However, each is distinguished by its ideological orientation, the former secular, the latter theocratic.
Sample Sources In this study three sampling strategies were employed depending on the source: (a) archival, (b) interviews, and (c) secondary analysis. I used mixed purposeful sampling for the three source categories.[2] Archival sources were the primary means of informing this study, including those available from governments, organizations, individuals, and the media. Furthermore, I conducted semi-structured and unstructured interviews with subjects living in Iran, Israel, the United States, Canada, and France. Finally, due to the scarcity of available and accessible information, secondary analysis sources became an important part of providing triangulation for other material and a means of providing information that was not accessed by the other two means.
Data Collection Archival Sources In dealing with a chronological case study spanning 85 years, archival information was the primary source of information for analysis. In the process of accessing governmental, organizational, individual, and media sources, and applying purposeful sampling, I created predetermined categories and subcategories for collection. Table A1[3] shows the general categorization and labeling scheme applied to sampling archival sources, which are described in greater detail below.
I sampled electronic sources in the above four categories by inputting key words into Internet search engines and word processing scanning applications. Because there were limited sources on the topic, I scanned for key words in every issue of those printed materials to which I had access. I used a hierarchical search technique, using primary key words, followed by secondary and tertiary key words input into search engines (details discussed below).
Purposeful and snowball sampling of printed sources was used by reading a wide range of secondary literature, identifying primary cited sources, and then seeking out those primary sources. Subsequently, I would search those primary sources for additional relevant information. Some sources were searched systematically and thoroughly, while others were accessed through general broad-based sampling using key word queries, employing an Internet search engine to locate additional sources. I was able to organize all material to provide sufficient information for triangulation and validation.
Archival instruments. In collecting the data, I used various technical instruments and skills, including translation, search engines, and research assistance. I used Persian and English sources for this study, drawing also on the work of a research assistant to expedite the reading of the Persian texts and identify relevant material. In looking for information sources and material, I developed a basic, flexible technique by applying hierarchical levels to key word sets in Internet search engines and word processing search features. I used a similar system for reading printed material.
Three general levels were used in search queries. Level one key words included case subjects and objects. Level two included primary themes of the study. Level three addressed specific issues, events, individuals and institutions, and other miscellaneous particulars arising from level two queries (varied and not systematic). These were then used in vertical and horizontal combination (e.g., Iranian Jews + education + Islamic Republic). Table A2 is a sample tabulation of the type of hierarchy of key word sets used to look for material directly related to this study. Tables A3 (Governmental Sources), A4 (Organizational and Individual Sources), and A5 (Media Sources) represent archival sources used in parts of this study. In addition to the four categories of archival sources mentioned earlier, the archival source tables specify the number of documents, types and location of sources, how they were accessed, and dates, when applicable.
Rationale. The sources selected for this study were chosen because, in addition to being useful in providing answers to the research questions, they were the most accessible to the researcher. They not only represent the kinds of sources that are generally used in studies of Iranian religious minorities, but in some cases include a wider range. Some sources which do not appear in the archival source tables in the Appendices, are found in the reference list, because they were not systematically retrieved, but rather resulted from unintentional referral or search. Other archival sources not included in these tables (or this study in general) are the result of limitations of the researcher (see section on limitations at the end of this chapter).
Interview Sources All three religious minority groups constitute vulnerable populations in Iran. It is not surprising that accessing information from members of any of the three groups was difficult, particularly because these groups have had to cope with restricted or risk-laden conditions by being circumspect in sharing information. Addressing the accessibility and availability of researching religious minorities in Iran, Eliz Sanasarian (2000), among the foremost experts in Iranian religious minority studies, asserts, “The scholarly literature on non-Muslim minorities is highly uneven, complex, and thinly researched” (p. 34). Knowing from preliminary research that access to human information sources would be challenging, and that information is sometimes distorted and generally guarded from out-group members, I set out to identify community leaders, organizers, and prominent members to be “key informants” for the study about community strategy selection. Key informants include individuals who were able to obtain “descriptive information that might be too difficult and time consuming to uncover through more structured data gathering techniques” (Blee & Taylor, 2002, p. 105). I initially employed purposeful stratified sampling by soliciting interviews from community religious leaders, organization representatives, and experts on each community, both inside and outside Iran. I defined the categories as follows:
1. Community religious leaders--rabbis (Jewish), clergy and pastors (Christian), Assembly members (i.e., NSA and LSA members, or equivalent for Baha’is)
2. Organization representatives--nongovernmental group organizations, community advocates participating in secular organizations, and committee members responsible for educational services and strategies
3. Experts--included academics and scholars who research one or more of Iran’s religious minority groups, nongovernmental and government agents who specialize in a particular group(s), lay members in the community who show significant knowledge about the group (i.e., journalists, rogue scholars, etc.)
I presumed that, without insider trust networks, I would not be able to effectively interview community leaders and members of groups inside Iran, and made contacts with individuals in Iran through contacts in the United States. Snowball sampling was also used in the interview process. In the course of my interviews with different group members (Jews, Christians, and Baha’is), I was referred to several members of the Iranian group communities who had experiences that were directly related to education in Iran. Through snowball sampling I was also able to access other community members who had experience with educational strategies used in Iran, including community run school administrators, teachers, students, and parents. Gender, age, and socioeconomic status were not included in sample design; only characteristics associated with roles in the community (i.e., community leaders, organization representatives, and prominent members) and educational experience in Iran was used in identifying subjects.
My initial target sample size was 10 community and organization leaders for each group, and three experts for each community; a total of 39 subject samples—given circumstantial restrictions I found this to be a sufficient number to provide essential information. With the goal of purposeful sampling of interview participants, I sent 19 independent requests for interviews with Jewish sources, 17 independent requests for interviews with Iranian Christian sources, and 14 requests for Baha’i sources (see Table A6). Contact information was obtained through Internet community websites, referrals by experts and other community leaders. These requests were sent by invitation letter (electronic) or in telephone calls. The acceptance rates were low. This was not surprising, given the high levels of tension in the social and political climate in Iran, and because of the sensitive nature of the topic and the understandable caution exercised by members of religious minority groups in general. In an attempt to offset the low response rates, I asked to be referred to others of similar rank; subsequently many pointed to the same source or directed me to an academic expert.
I was informed by several contacts that conducting research in Iran would not produce substantial results because of the “closed” environment. I was also warned that I should exercise caution because of the topic, as well as my own affiliation with an unrecognized religious minority group (i.e., Baha’i). However, after careful consideration, I decided— because of the dearth of existing information about Iranian religious minorities and educational issues —that it was necessary for me to travel to Iran to conduct anonymous, unstructured interviews. In Iran, I conducted interviews based solely on snowball sampling which began outside Iran through referrals, or while I was in Iran. I used extreme caution in interviewing members of Iranian religious minorities, not only because they are vulnerable, but also because of the sensitive nature of the dissertation topic.
Initially, I had set out to conduct five unstructured interviews with Iran-based informants involved in education of religious minorities for each group (i.e., students, teachers, administrators, etc.). However, my access to the minority school personnel I contacted—who will remain unidentified— was categorically denied. This clearly meant that approaching minority-run schools in the context of the current political climate in Iran was safe for neither the interviewee nor the interviewer. I was able to interview 28 Baha’is in Iran because of the insider sources I had already established prior to arrival in the country, but primarily because of the additional time that resulted when I was unable to access Jewish and Christian interviewees. I was able to interview two Christians and three Jews living in Iran. Most of those whom I interviewed wished to remain anonymous. Tables A6 and A7 reflect the sampling numbers by purposeful sampling and snowball sampling (based on an extensive interview database that was created to keep track of interviews by group, organization affiliation, position/role, location, type of communication, and dates).
Interview instruments. An interview guide was prepared and approved by the Institutional Review Board (IRB) at Teachers College, Columbia University, to be used in interviews with community and organization leaders, prominent community members, and topic experts (Table A8 shows the interview guide). The guide was nominally adjusted depending on the individual’s role and group affiliation, and translated into Persian. The interview guide included questions that solicited descriptive and explanatory answers related to their group and educational strategies. The guide included questions related to each group’s educational opportunities and challenges in Iran (past and current), as well as the group’s characteristics and composition, networks, and regime relations bearing on educational opportunities and strategies. I remained flexible and adjusted the interview guide according to each participant’s receptivity and expertise with the topic. I also left time for the participants to share information they felt would inform the study. This was particularly the case with those who did not know much about educational issues in the community, but were knowledgeable about other matters bearing on education, such as regime-group relations and community characteristics and composition.
I also employed unstructured interviews for follow-up session with interviewees, as well as all Iran-based participants. The usefulness of unstructured interviews was apparent when participants shared information about issues which did not appear in any other primary or secondary source, or about which I had no previous knowledge. I incorporated several themes in unstructured interviews, particularly questions related to motives, fears, considerations in strategy selection, as well as experiences of opportunity and challenges related to educational pursuits. In addition, at some point during the unstructured part of the interview, I asked about other individuals whom I could contact. I also included written communication as part of the unstructured interview process, usually in the form of two or three questions soliciting answers about a particular subject, source, or fact. These were equally important, especially when retrieved from a key informant.
Interviews were conducted in person and using communication devices, in both English and Persian. In most interviews outside Iran, I used a voice recorder and took written notes. In Iran, I took only encrypted notes, and would then write out and email expanded notes to myself (still using cryptic language). I kept a ongoing database, into which all interviews were logged by participant descriptor when available (i.e., gender, age, initials to remind me who it was), date, position/occupation, organization/notes, date of interview, modality (i.e., written correspondence, personal communication, or telecommunication), and location.
Interviewees reside in various cities around the world, including Canada, France, Iran, Israel, and the United States. I spent 2008–2010 in Los Angeles, California as my main location of field research, because most members of Iranian religious minorities living outside Iran live in Southern California—particularly Jews, Armenian-Iranian Christians, and Baha’is. I traveled to Iran in December 2009 to conduct interviews. The political climate in Iran during this period was unstable as a result of the post-election protests and riots of 2009.[4] However, I was able to travel and conduct interviews in Tehran, Shiraz, Babol, and Sari during my stay.
Consent and Confidentiality Given the vulnerability of group members and the sensitive nature of the study topic, I extended anonymity rights to all those whom I interviewed (unstructured and semi-structured). Most interviewees chose to remain anonymous. I followed IRB regulations for semi-structured interviews. I obtained oral consent and verbally conveyed stipulations of confidentiality with interview participants to reduce potential risk to participants in the study. The consent form included the purpose of the study, the researcher’s background (personal and academic information, and contact information), and confidentiality particulars. All participants were specifically informed that they could stop the interview at anytime, and did not have to answer any questions about which they felt uncomfortable.
Secondary Analysis and Sources Secondary analysis and literature was an important part of this study, as it provided important background information on settings, regimes, and subjects. The information regarding religious minorities in Iran is scattered, fragmented, and sometimes unreliable (Sanasarian, 2000). One contribution which this dissertation makes to the study of religious minority groups lies in the efforts to synthesize and critically analyze the existing primary and secondary information in this field. I used several academic secondary analyses, cited throughout, to complete a comprehensive narrative of otherwise fragmented streams and episodes describing educational experiences of religious minority groups. The use of primary sources and secondary analysis also reinforces triangulation of validity of data (see validity issues in section below). I attempted to access as many secondary analyses of Iranian religious minority community issues related to education. Much of the secondary source material gathered for all three groups repeats information and often cites other similar sources. This was useful in identifying the limits and strengths of secondary literature.
Data Analysis Methods Theoretical Propositions I used preconfigured theoretical propositions to categorize collected data.[5] The study follows a historical timeline with distinct actors in question. Thus, I arranged the information along these two dimensions. The two major categories/themes are regimes and religious minorities, in order to optimally organize the data for a chronological multicase study, using the mechanism-process approach. Three other preconfigured codes were also assigned to religious minorities, intrinsically related to the research question: (a) group composition and characteristics; (b) group networks; and (c) group-regime relations. Finally, the category education was used to identify information directly related to educational opportunities, challenges, and strategies in modern Iran. Figure 4 illustrates the basic organization of data by category. The categories and labels were used to organize and manage collected data from archival sources, interviews, and secondary analysis.
Figure 4. Preconfigured proposition for categorizing data.
Interviews Analysis After collecting recordings or notes from interviews, I scanned the content and used the preconfigured thematic codes (see Figure 4) to label interview content. After listening to the recorded interviews, information was included which was either missing or incomplete in my corresponding interviews notes. When information was shared that was outside the bounds of the categories, I would measure its perceived importance by how it would contribute to the study and labelled it “miscellaneous (short description).” Not all content derived from interviews was of equal length or quality. However, because of this labeling scheme (dividing up interviews by religious minority group), I was able to access interview content easily when looking at other data for the given group and time period.
Four distinct categories emerged in the data collected in interviews:
1. Historical--events (episodes, interactions, etc.)
2. Descriptive--community characteristics, statistics
3. Relational--relations with the government and other religious minorities; networks within and outside the community
4. Referential--additional potential people to interview, and additional sources
Most interviews were used in the process of constructing historical events for each period, but also to inform particulars about how characteristics, networks, and regime relations affected educational and other social strategies. In many cases information was supplemented by archival and secondary sources, but also used to supplement archival sources that were incomplete. Interviews added a rich texture to the study, particular for the Islamic Republic period, during which so little has been written about Iranian religious minorities.
It would be appropriate to mention here that I used secondary interviews in my study, but treated them as archival sources. I used a similar process of labeling those interviews using my theoretical propositions. Of particular value were the Center for Iranian Jewish Oral History and the Foundation for Iranian Studies Oral History Program. Short interviews available in documentary films contributed additional empirical evidence for the study.
Historical Analysis Historical inquiry. Using preconfigured codes listed in Figure 4, I set out to find a substantial body of primary sources to outline the historical development of educational processes which could then be broken down into bound streams and episodes (see mechanism-process approach below). I employed historical inquiry and interpretation as my initial means of analyzing and organizing this information. Historical inquiry is not just telling a story. Rather, it is the critical analysis of sources from various periods in the past, interpreting events and interactions, arranging them in logical sequence, with the goal of explaining what, how, and why a phenomenon in question occurred in the past. The categorized material was then systematically organized so as to be readily accessible when I moved through a chronological sequencing of events relating to educational development in Iran, for the religious minority groups specifically. Part of the process of analysis was to determine the validity and quality of the collected sources. In most cases, I was able to triangulate the data with other primary or secondary sources. However, in others cases, the information gathered was all that was available to me, and I had to use discretion in relying on those sources. In general, depending on the type of source, I used various preanalytical methods in drawing out useful information to create a historical narrative, against which I could apply the mechanism-process approach.
Document and media analysis. Concepts from document analysis were identified by themes in the texts. Those (particularly media) which contained the religious minority titles (i.e., coded as any derivation of Jew, Christian, Armenian, Assyrian, Chaldean, and Baha’i) were classified using a filing system. For print media (Ettelaat, Iran Times and Kayhan), I examined all issues for these primary codes. For raw archive material, such as letters, reports, and pictures, I collected information using the same classification scheme as in Figure 4, noted the content that related to the study, and filed it under the appropriate category. I was able to effectively examine how group composition and characteristics, networks, and regime-group relations affected strategy selection by using historical inquiry and interpretation of events and interactions within this designated framework. I did not engage in a quantitative analysis of the documents nor do an in-depth discourse analysis of the text. The purpose of the document analysis was to construct a holistic and balanced historical narrative which could be further analyzed through the mechanism-process approach. It should be noted that in Chapter 5, I discuss the particular features of group characteristics and composition, networks, and regime-group relations, including definitions of concepts.
Mechanism-Process Analysis This study set out to explain how a group meets its educational needs. While I discussed the theoretical background of the mechanism-process approach at length in the literature review, it is important to consider how it applies directly to this dissertation. I employed the mechanism-process approach in Chapter 6 in order to identify specific streams and episodes of interaction that significantly influenced educational strategies and their outcome. In the process of historical analysis using various sources, I specifically looked for streams and episodes of contention and actuation. As Tilly and Tarrow (2007) describe, “episodes are bounded sequences of continuous interaction, usually produced by an investigator’s chopping up longer streams of contention into segments for purposes of systematic observation, comparison, and explanation” (p. 36). I also use the terms streams of actuation and episodes of actuation to denote regime-group interactions characterized by facilitation and tolerance. Although more unwieldy than contentious interactions, streams and episodes of actuation highlight important periods of strategy development for the case subjects when conditions and government education policies are more favorable than contentious. Based on available data for each period (i.e., Pahlavi and Islamic Republic), streams and episodes related to education opportunities and challenges become the units of analysis in studying educational strategy selection. I identified episodes of contention and actuation for the three religious minority groups. Often, several smaller events or micro-episodes were identified and analyzed within episodes. Once episodes were identified and described, I proceeded to apply the eight steps in the mechanism-process approach as outlined by Tilly and Tarrow (2007, p. 207, summarized in Table A9).
The purpose of analyzing my historical interpretation through the mechanism-process approach was to explain a) how strategies were actually selected, developed, and deployed, and b) how such strategies affected the subsequent selection of other strategies in meeting educational needs. In Chapter 2 (Literature Review), I defined essential terms such as mechanisms, processes, and episodes. Due to variation in the extent of information available and accessible for each time period and across the different groups, including general differences in group features, multiple scales of observation were used to identify processes. To identify mechanisms and processes over time, I borrowed the categories developed by Tilly and Tarrow (2007), and interpreted strategies in those terms (see Table A10 for a listing and definitions of mechanisms and processes used to explain strategy selection during analysis in Chapter 6). I also include specially created content-specific processes when analyzing the episodes, and describe these in greater detail in Chapter 6. While processes were identified and reassembled to explain the selection and development of strategies, I also interjected an interpretation of how each group’s composition and characteristics, networks, and regime-group relations influenced the shaping of those processes.
Concepts of Network Analysis In Chapter 5, concerning the religious minority groups’ composition and characteristics, networks, and regime-group relations, I used concepts (not mechanics) of network analysis to visualize the domestic and international network ties of each group. Instead of using specifically designed surveys or computation of archival data—input into a network analysis software program, as typically used for network analyses—I made a simplified and holistic estimate of network ties and their values, based on a range of qualitative and quantitative sources available to this author.
Five nodes were classified for inclusion in the network mapping. For each group, I looked at following prominent actors:
1. Local community organizations and members
2. Transnational community organizations and members
3. Nongovernmental organizations (nongroup)
4. Governments and their agencies
5. International and supranational organizations (nongovernmental and governmental institutions)
I initially identified node ties with the subject group by labeling archival sources, interviews, and secondary literature that mentioned such connections with a code for network tie (i.e., NT). Nodes or actors were identified by geographical location. Since technical network analysis methods are beyond the scope of this dissertation, I only use approximated measures to determine the values of network ties. The sources I use to determine network nodes, tie strength, tie relation, and tie type, include government and organization documents and sources, organizational source material, news media sources, in-depth interviews with organizational leaders (15 participants inside and outside Iran), and other archival and secondary sources (see Tables A2, A3, and A4 for the list of sources used). Four features are included in my conceptual network mapping exercise: network nodes (i.e., actors); types of ties between nodes; relational content of ties; and strength of ties between nodes.
I narrowed the focus on three other basic features as part of my conceptual network analysis:
1. Types of relational ties between nodes—direct, indirect, independent
2. Relational content of ties—information, resources, advocacy
3. Strength of ties—between nodes
Three types of relational ties are identified in my approximation. Direct ties indicate direct exchange or interaction between an Iranian group community and another node (whether in or outside Iran). An indirect tie indicates that one of the three interactions (informal, advocacy, or resources) exists, but through an additional party (i.e., broker) working as a conduit of exchange. The term independent ties indicates existing network ties between different entities (particularly governments), representing state-state relations or organization-organization cooperation. Drawing on international relations theory, I argue that these relations have an indirect impact on minority groups who are associated with communities in those particular countries.
Examined here are those ties involving the exchange (unidirectional or bidirectional) of information, advocacy, or resources.[6] Information refers to that which can be utilized for reports and assessments, not information in the sense of resources used toward community development (e.g., curriculum or school models). I use the term resources to refer to the categories of Edwards and McCarthy (2004; see literature review), including material, moral, socio-organizational, human, and cultural. Advocacy refers to any form of initiative that denotes explicit protest or support of particular regime policies and practices, or group actions; some forms of advocacy include mutual agreements, petitions, letters of appeal, voiced concern, motions made in multilateral supranational organizations, and sanctions. In the network maps in Chapter 5, icons at the end of the connecting lines identify content ties (information, resources, and advocacy).
To estimate the measure of tie strength, I use the descriptive values of weak (s = 1), moderate (s = 2), strong (s = 3). It is especially important to emphasize that network strength was estimated by accounting for recorded frequency of interaction, level of actions (i.e., prominence), and longevity of the relationship. Strength is not computed scientifically, but is attributed to emerging trends that surface from review of primary and secondary sources. In the network maps in Chapter 5, strength is represented by the thickness of the lines connecting nodes.
Validity Issues Several research methodologists have suggested that to reduce threats to the validity of a given study, triangulation of data collection ought to be incorporated into the study (Creswell, 2007; Maxwell, 1996; Yin, 2009). In this study I used archival sources of various types (media journals, letters and memoirs, organization documents and publications, government documents and publications), semistructured and unstructured interviews, and secondary analysis to provide adequate triangulation.
Nevertheless, some consideration of potential threats to validity is warranted due to general research bias, interview participant bias, and source bias. My background and affiliation with an Iranian religious minority group, could be considered a potential threat to the study. Growing up in a community that framed the situation facing the Baha’is in grave terms created a preconception of the situation facing religious minorities. This was offset in a number of ways. By adopting the idea that all regimes engage in repression, toleration, and facilitation, I moved out of the fallacious dichotomous approach to government behavior (i.e., repressive or nonrepressive). More importantly, in using Baha’i sources or interviews, I exercised extreme discretion by triangulating data from non-Baha’i sources (i.e., governmental and nongovernmental, and more specifically from academic expert analyses). Conducting interviews with vulnerable religious minorities also poses some challenges. Validity of information gathered from these interview participants could have also been compromised when participants were fearful of having information used against them or others. Here again, I relied on triangulation to confirm information, using other interview sources, archival sources, and secondary analysis. Additionally, key informants were extremely helpful in compensating for the low number of interviews. Finally, there are sources on Iranian religious minorities (whether primary or secondary), and some are clearly biased. Thus, lack of information may also have given rise to nuances that were not accounted for in this study. However, I am confident that through the use of multiple sources, major threats to validity were overcome.
Limitations The limitations to this study can be categorized as those related to language, access, safety, and current climate in Iran. I used Persian and English to conduct the study, but was unable to use sources in Hebrew, French, Armenian, and Syriac. As mentioned earlier, conducting research on religious minorities in Iran is an extremely difficult task because of the sensitivity of the situation and the reticence of these groups since the Islamic Revolution. The caution and insularity of Iranian religious minority communities has only been amplified by current political and social restrictions during the presidency of President Ahmadinejad (2004–2009), the increase in religious minority harassment and slander in media, and the post-election protests and government crackdown since 2009. Thus, compromised trust and restricted conditions by the government inhibited freer access to more sources. The need for constant vigilance and concerns for safety were other limitations. I did not want to put my interviewees or myself at risk, particularly those participants who live or have connections in Iran. Similarly, I was advised by group informants that Christians and Jews are extremely insular in Iran, and even those who have recently left are reluctant to share information with outsiders. Notwithstanding these limitations, by using a wide range and differing types of sources, I was able to collect sufficient information to undertake this essentially unstudied research.
[1] I did not include the Zoroastrian community because of small numbers, limited access to sources, and time constraints; i.e., there are only some 10,000 Zoroastrians in Iran, living in mid-size cities and rural areas (Sanasarian, 2000).
[2] According to Johnson and Christenson (2008), “mixed purposeful sampling is likely to be used when a researcher uses data triangulation—examining multiple data sources which might be selected according to different sampling methods” (p. 246).
[3] See Appendix A for all Tables in this chapter.
[4] Two days before my arrival there had been mass protests in Tehran, and on the day of my departure there was an escalation of violence in Tehran.
[5] With regard to case studies, Yin (2009) writes: “The first and most preferred strategy is to follow the theoretical propositions that led to your case study. The original objectives and design of the case study presumably were based on such propositions, which in turn reflected a set of research questions, reviews of the literature, and new hypotheses or propositions. The propositions would have shaped your data collection plan and therefore would have given priorities to the relevant analytical strategies” (p. 130).
[6] There may be other dynamic interactions present, but for the sake of brevity and this heuristic exercise, I have retained focus on only these three features.