Christian Relief, Development, and Advocacy 1(2), Winter 2020 Kiš, Cabus, Romo, and Welch, “We are poor, but we are developed” 93 “We are poor, but we are developed:” Differential Perceptions of “Development” among Key Stakeholder Groups in the Philippines Adam D. Kiš, Jovy Conde Cabus, Tzaddi Romo, and Cheyanne Welch This mixed-methods study conducted in the Philippines explores the perceptions of various stakeholders in the international development process on the meaning of the word “development.” It also measured the extent of differences between stakeholder groups (Adventist Development and Relief Agency [ADRA] employees, government liaisons, and beneficiaries) on their perceptions of what “development” means. Results show that approximately half of the self-generated indicators associated with the word “development” overlap between stakeholder groups, and approximately half are unique to each stakeholder group. In addition, the differences measured between stakeholder perceptions are statistically significant using ANOVA and MANOVA procedures. The implications for these different perceptions according to stakeholder group are discussed. Furthermore, the implications for faith-based NGOs working in development (such as ADRA) are also discussed. Introduction and Literature Review “Development.” This single word has inspired countless humanitarian efforts, spawned a multibillion- dollar industry, and impacted every country in the world. One might expect such a powerful term to have a universal definition; alas, it does not (Heryanto and Lutz 1988; Arndt 1981). One of the major divisions is whether to view it from a macro or micro perspective; is development about GDP growth and the institutions that support (or hinder) it, or is it about small agencies and local governments addressing the specific needs of the poor? Much theoretical discussion has taken place around these questions, but ultimately “development” is an “essentially contested concept” (Gallie 1956). In addition to these theoretical discussions, abundant anecdotal evidence repeatedly affirms that the concept of “development” is not understood in the same way by everyone. To show this conclusively, rigorous academic research is required. Even if perceptions differ between individuals, there may be some commonality between groupings of people that share certain characteristics. What is salient when it comes to defining the term “development,” is 1) whether there is cohesive perception within stakeholder groups involved in the development process, 2) whether there is commonality between stakeholder groups, and 3) what the implications of any differences between stakeholder groups might be for long-term outcomes of development interventions. This study focuses primarily on the second question, while touching upon the first and third. A fair amount of literature addresses similar – though not identical – questions. We have organized that literature into classes, each one progressively fulfilling more criteria that we deem important for a rigorous empirical study on the meaning of “development” to have (these are the same criteria upon which we designed our own study). For example, studies should ideally address the meaning of “development” directly, and not simply through associated terms. Ideal studies should also compare at least two perspectives on what “development” means. We are especially interested in studies that include beneficiary perspectives, and not just the perspectives of those on the giving end of development assistance. Finally, it is important that studies that meet all of the previous criteria also be empirically-based rather than reason-based; we sought studies that were field-tested, and not just conceived through rational means. These search criteria are summed up in Table 1. Christian Relief, Development, and Advocacy 1(2), Winter 2020 Kiš, Cabus, Romo, and Welch, “We are poor, but we are developed” 94 C la ss 1 C la ss 2 C la ss 3 C la ss 4 C la ss 5 C la ss 6 Defines “Develop- ment” X X X X Includes Multiple Perspectives X X X Includes Beneficiary Perspectives X X X Field Researched X Table 1: Literature Search Criteria and Classes of Studies Found Class 1 studies are those that do not directly address the definition of “development” itself, but rather explore concepts related to development. Additionally, these studies did not consider multiple perspectives and they are not field researched. Nevertheless, they are included in the literature review because they highlight the substantial scholarly activity around the semantics of key development terms. Studying the meaning of development terms is clearly not a fringe exercise. Examples include Cornwall and Brock (2005), who consider how “participation,” “empowerment,” and “poverty reduction” have morphed over time to mean very different things than before, and how this change in meaning impacts those terms’ ability to engender development. Holmberg and Sandbrook (1992) argue that “sustainable development” needs to be more precisely defined. Tipple and Speak (2005) advocate for defining “homelessness” in situ so that homeless people can be accurately identified and targeted for development assistance. Eyben and Napier-Moore (2009) discuss how the meaning of “women’s empowerment” within the context of international development has many shifting meanings, depending upon who is wielding it and what their intentions are. And Campbell and Vainio-Mattila (2003) posit that “participatory development” and “community-based conservation” are related terms, and advocate for further exploration and clearer definition of those terms as a means of making them function synergistically. 1 The Organization for Economic Co-operation and Development’s Development Assistance Committee. The OECD-DAC serves as a forum for the world’s most developed countries to discuss and coordinate aid strategy. There is much overlap between G7 and OECD-DAC membership, although OECD-DAC membership is considerably larger. Being a member of OECD-DAC is a tacit acknowledgement by other members that they consider the country to be fully developed, implying that its only role in development intervention should be that of an aid donor and not a Class 2 studies continue in the same vein as Class 1, but also consider the perspectives of beneficiaries. Examples include Marcus (2001), in which the impact of conservation projects in Madagascar upon beneficiaries’ perceptions of the meaning of “conservation” are investigated. Lipi (2016) explores the meaning of “dignity” and “empowerment” to beneficiaries of a women’s microcredit program in Bangladesh. And Quinn et al. (2003) assesses local perceptions of “environmental risk” in Tanzania. Class 3 studies begin to address the meaning of “development” itself. Seers (2010), for example, defines “development” in the context of government priorities. He writes, “The questions to ask about a country’s development are therefore: What has been happening to poverty? What has been happening to unemployment? What has been happening to inequality? If all three of these have become less severe, then beyond doubt this has been a period of development for the country concerned” (Seers 2010, 12). Ingham (1993) analyzed how the meaning of “development” has shifted in response to the spirit of the times (people-centered development, sustainability, and so forth). Yet all of her examples focus exclusively on the perceptions of those on the giving end of development assistance. Blowfield (2005) reports on how some have redefined “development” as Corporate Social Responsibility (CSR), and suggests that such a redefinition may not be warranted. Offutt and Reynolds (2019) look at how “development” is defined by leaders of Faith-Based Organizations (FBOs), finding that many subscribed to the transformation paradigm: development as restoring broken relationships (human to human, human to God, etc.). Diallo and Thuillier (2004) discuss the success of “development” as defined by the perceptions of development project coordinators in Africa. And Sen (1999) elaborate upon “development as freedom” in a book by that title. Class 4 studies take the focus on “development” of Class 3 studies and include multiple perspectives. Examples of this include Horn and Grugel’s study which reports that the Sustainable Development Goals (SDG’s) are understood and prioritized differently according to stakeholder groups in Ecuador (Horn and Grugel 2018). In this particular case, the two stakeholder groups under comparison are national government and Quito city government officials. And a cluster of studies all pointed out the differences between OECD-DAC 1 and non-OECD-DAC donors Christian Relief, Development, and Advocacy 1(2), Winter 2020 Kiš, Cabus, Romo, and Welch, “We are poor, but we are developed” 95 on what “development” means, highlighting how this can get in the way of streamlining cooperation and harmonizing work toward a common goal (Zimmerman and Smith 2011; Bräutigam 2011; Chandy and Kharas 2011). Class 5 studies move beyond simply including multiple perspectives to specifically including beneficiary definitions. For example, Cavalcanti (2007) reports on the conflicting views of what “development” means to a development agency and its beneficiaries in Brazil. Cavalcanti do not, however, explore those perceptions directly with beneficiaries; instead, he speaks with development technicians and reports their negative opinions of beneficiaries for not getting on board with the project. Bell and Morse (2011) reveal how indicator selection for “development” can be tyrannical if exercised by powerful external agencies upon beneficiaries who don’t contribute to that definition. Their article strongly advocates on behalf of the inclusion of beneficiary voices, but is not itself based upon field research with beneficiaries. Finally, Scheld (2018) discusses how the difference between NGOs and the common people in defining Haiti’s development trajectory ultimately holds back Haiti’s development. Class 6 studies meet all of the criteria specified by Table 1: they focus on defining “development,” they include multiple perspectives, they specifically include beneficiary perspectives, and they are based upon field research. Parkinson (2009) presents a case study of a development program in Uganda, describing a mismatch between the assumptions of the government implementers and beneficiaries on what the goal of “development” is for the project. The impacts of this mismatch are analyzed in light of Participatory Monitoring and Evaluation (PM&E), with the finding that beneficiaries do not buy into program implementers’ visions and feel that bureaucratic and accountability requirements are not their responsibility. Carroll-Bell (2015) analyzes three case studies of NGOs in Timor-Leste that explore strategies for development that take local definitions into account. Citing Lewis and Mosse (2006:9), Carroll-Bell writes, “the authors argue that significant ethnographic research is needed to elucidate ‘the ways in which development meanings are produced and negotiated in practice and how development processes and interactions have different significance for the various actors involved’” (2015:314). The present study is one such research project, though it differs from Parkinson (2009) and Carroll-Bell (2015) in its methodological approach, avoiding the case study method in favor of an ethnographic survey. recipient. Nevertheless, some countries play the role of donor, despite not being considered developed enough to merit OECD-DAC membership (and, of course, many countries play only the role of aid recipients). It should not be construed that any of the foregoing literature that did not meet all of the criteria in Table 1 is somehow deficient or inaccurate. The criteria in Table 1 were merely used to highlight the characteristics of the research study that we aimed to do, and to demonstrate that many of the previous studies did not take the particular approach that we deemed necessary to fill a specific gap in the literature. Background All of the foregoing contested meanings of “development” and its associated terms have an impact on the day-to-day operations of development agencies at all levels. Even if large institutions don’t pay much attention to differences in perception at the grassroots level, differences of perception can still impact the rollout of development plans and programs. Thus, this study was designed around one particular development agency and its associated stakeholders in order to understand the nature of perceptions on “development” at various levels, to highlight areas where those perceptions may differ, and to propose ways in which differential perceptions can be transformed from barriers into opportunities. The Adventist Development and Relief Agency (ADRA), with headquarters in the United States, is a faith-based international NGO that works in the domain of humanitarian relief and international development in more than 120 countries around the world. It is the official humanitarian arm of the Seventh- day Adventist Church. Although it is motivated by love for God and concern for fellow humans, ADRA does not discriminate in its programs; one does not have to be a church member in order to benefit from ADRA’s projects, and church membership is not even a criterion for beneficiary selection. Furthermore, ADRA does not discriminate by religion in its hiring practices, save for the most senior positions which do require membership in the Seventh-day Adventist church. Finally, ADRA is explicitly non-proselytizing; there is no expectation of conversion on the part of beneficiaries, and such conversion is neither actively sought nor explicitly discouraged. Though owned and operated by the Seventh-day Adventist church, ADRA exists to serve the needs of all humanity regardless of religious persuasion. To explore perceptions of the meaning of “development” by various stakeholder groups in the development process, we undertook a research project in Leyte province, Philippines in 2016. Although the Principal Investigator (PI) never worked for ADRA in the Philippines, he did work for ADRA in sub-Saharan Christian Relief, Development, and Advocacy 1(2), Winter 2020 Kiš, Cabus, Romo, and Welch, “We are poor, but we are developed” 96 Africa for six years. Having lived in the Philippines for three years himself, the PI’s personal acquaintance with the management team, and ADRA Philippines’ openness to research s helped the PI select the Philippines as a research site. ADRA Philippines provided logistical support for various aspects of the research project, but they neither commissioned it nor financed it. The research project was born entirely of the PI’s initiative and was completely paid for through research funding from the PI’s North American academic institution. This minimized research influence from ADRA Philippines while providing access to ADRA Philippines’ personnel and project sites. In consultation with the ADRA Philippines team, Leyte province was identified as a prime location to conduct this study. ADRA was heavily involved in disaster relief projects following Typhoon Haiyan (known locally as Yolanda) which ravaged the province in 2013. Over time, ADRA’s interventions transitioned to long-term development projects as the province was restored to previous operating levels by both national and international actors. As the majority of humanitarian agencies moved on to other disaster zones around the world, ADRA stayed behind, aiming to improve the quality of life for Leyte residents beyond pre-Haiyan levels. This means that ADRA has had a significant presence in the province – among the criteria deemed important for a study such as this. At the time of this study, ADRA Philippines worked primarily in two regions of Leyte province. The province occupies the northern three-quarters of Leyte island, which is located on the east-central side of the Philippine archipelago, facing the Pacific Ocean. In the city of Tacloban and its environs (located in northeast Leyte), ADRA operated the ERL project (Economic Recovery in Leyte), a livelihoods project that aimed to restore (and possibly supersede) previous levels of economic self-sufficiency for residents impacted by the storm. In the municipality of Bato and its environs (located in southwest Leyte), ADRA operated the MASIGLAHI maternal and child health project (an acronym in the local language that means “Better Nutrition for Mothers and Children”), which “aims to reduce acute and chronic malnutrition among children within their first 1,000 days to prevent irreversible effects of malnutrition” (http://adra.ph/index.php/causes/masiglahi/). It is in these two project regions that the current research study was carried out. Methodology This research study used a mixed-methods design. The initial qualitative phase took place in Bato and its environs as a way to understand the nature of development stakeholders’ perceptions. Open-ended methods were used to elicit descriptions of what people perceive when they hear the word “development.” From those descriptions, quantitative surveys were developed and administered in Tacloban and its environs in order to determine the extent of differing perceptions among the various stakeholder groups. The hypotheses tested were as follows: H0: There is no significant difference between stakeholder groups on perceptions of the meaning of “development.” H1: There is a significant difference between stakeholder groups on perceptions of the meaning of “development.” Creswell and Clark (2006) describe a number of different models for mixed-method studies. Some studies are primarily quantitative with the qualitative portion serving a supplementary role; others are primarily qualitative with the quantitative portion serving a supplementary role; and still others are equally weighted, with qualitative and quantitative portions serving complementary roles. This study falls under the first type: a study in which the primary aim is to measure quantitative differences between stakeholder groups in their perceptions of “development,” but where the qualitative phase was a necessary preliminary step which enabled better quantitative measurement. Creswell and Clark define this as an exploratory sequential design where the qualitative phase provides data upon which to build and test the quantitative phase. They write that this method is appropriate when one wants to “explore a phenomenon in depth and then measure its prevalence” (Creswell and Clark 2006, 75). Since data collection occurred sequentially (qualitative first, then quantitative) and because development of the quantitative survey was dependent upon the results of the qualitative phase, both methods and results are presented in the same order in which they took place during data collection: qualitative methods then qualitative results, quantitative methods then quantitative results. But first, a word about the languages of data collection and how that impacted understanding of the word “development” for both stages. All interactions with ADRA personnel were conducted in English. Knowledge of English is a prerequisite for most jobs at ADRA, and since English is one of the official languages of the Philippines and is widely taught in schools, anyone with a substantial amount of education speaks English with a fair amount of fluency. The language in which most qualitative interviews and focus group discussions in the field were conducted, however, was Cebuano, the local language in southwest Leyte, and most quantitative surveys were conducted in Waray-Waray, the local language in northern Leyte. Christian Relief, Development, and Advocacy 1(2), Winter 2020 Kiš, Cabus, Romo, and Welch, “We are poor, but we are developed” 97 This is because beneficiaries have substantially less fluency in English than do ADRA employees. This raises an important concern: different languages were used to conduct a study on the conception of a single word, “development.” By the very act of translation, “development” becomes another word – one that has an infinite range of possibilities for correlation with the precise meaning in another language. Still, the extent to which the inexactitude of translation may have affected our results in no way surpasses the extent to which the same inexactitude of translation impacts the sharing of a common perception by all stakeholders in the international development process. In other words, whatever challenges this introduced into the validity of our study were exactly mirrored by the challenges all development organizations face when working across a language barrier. That is precisely part of the point: beneficiaries may not understand the purpose and intent of development interventions in the same way as implementing actors partly due to linguistic difficulties. By entering that same murky netherworld of linguistic imprecision that ADRA Philippines and other international development agencies dwell in on a constant basis, we could conduct a sort of participant- observation on how this affects communication and the sharing of perceptions and ideals pertaining to “development.” One might wonder why we chose to enter into this linguistic quagmire in the first place; why not interview all Filipino study participants in their local language? The reason is that Cebuano and Waray-Waray are only two of many languages spoken in the Philippines, and they are not spoken by many ADRA employees, who hail from all regions of the Philippines and speak a variety of mother tongues. Thus, conducting research in English, Cebuano, and Waray-Waray was the minimum number of languages possible to accomplish this study as designed. Qualitative Methods In order to generate descriptions of what development looks like in each stakeholder’s mind, semi-structured interviews and focus group discussions were held with respondents from each of three stakeholder groups: 1. ADRA Philippines employees; 2. Local government officials who directly liaised with ADRA projects; and 2 A barangay is a local government unit below the level of province, akin to counties or townships in North America. 3 Although barangay health workers are technically government employees, these respondents were also beneficiaries of the ADRA project, and so were included in the focus group. 3. Development beneficiaries in communities where ADRA was working. Interviews of beneficiaries and government officials took place within the communities where ADRA was working, while interviews of ADRA employees took place within ADRA offices. Eleven ADRA employees, one government official, and eighteen beneficiaries were interviewed one-on-one. Across the thirty individuals interviewed, twenty were female and ten were male. Professions of respondents varied from positions within ADRA to occupations within the communities, such as nutritionist, farmer, fisherman, nurse, housewife, and small business owner. The average years of education for beneficiaries was 7.8 years, for ADRA employees 15.2 years, and for all respondents 10.7 years. Of particular concern was the low number of representatives from the local government. It was difficult to find them in the communities where ADRA was working because they frequently circulated within their jurisdiction, some of which fell outside ADRA project areas. Thus, a focus group was convened in conjunction with a regularly-scheduled meeting of government officials as a way to boost the number of respondents from this category. This focus group consisted of seven individuals, all female. The professions within this group included barangay councilors 2 , government-employed nurses, and experts on agriculture and nutrition. The average years of education were 11.9. Finally, although the overall number of beneficiaries who participated in individual interviews was adequate for our purposes, we conducted a focus group discussion with beneficiaries. The format of a focus group is such that it can elicit different kinds of information than would be generated during an interview. Consequently, we convened a focus group of 10 beneficiaries in order to discuss what they perceive “development” to mean. All participants were female. The majority identified as housewives, while the others were seamstresses, single parents, and barangay health workers 3 . This group averaged 9.8 years of education. Given that the interviews and focus group discussions were semi-structured, all respondents were asked the same initial questions, but were allowed to diverge as necessary in order to give a more complete picture of their perceptions. Research assistants were also allowed to ask follow-up questions that were not part of the initial list in order to explore answers further. The starter questions were as follows: Christian Relief, Development, and Advocacy 1(2), Winter 2020 Kiš, Cabus, Romo, and Welch, “We are poor, but we are developed” 98 1. When you hear the word “development,” what images/descriptions come to your mind? 2. Describe for me what you think a developed country or society looks like. 3. Describe for me what you think a developed community looks like. 4. Describe for me what you think a developed person looks like. 5. Do you consider yourself to be a developed person? Why, or why not? The first question was intended to elicit free lists of association with the specific term “development.” It was also meant to get the respondents’ minds thinking about development generally before delving into more and more specific questions. Questions 2 through 4 asked respondents to think about development at three successively narrower levels: large (country or society), medium (community), and small (individual). Once the focus was on the individual, it set the stage for question 5, which asked about development on a personal level. Qualitative Results The primary purpose of the qualitative phase was to develop a list of terms that could be used to describe the meaning of “development” to various stakeholders. We were also interested in illustrative quotes that could expound upon those short descriptors. But in terms of qualitative analysis, we initially restricted ourselves to extracting specific terms from interviews and focus group discussions and then merging terms that were synonymic. For example, if one respondent said “cell phones” and another said “mobile phones”, we would consider that an identical response. We were not creating categories so much as standardizing terms. That being said, some respondents listed discrete items (i.e. “cell phones”) whereas others spoke in categories (i.e. “technology”). Thus, our list of terms derived from respondents includes a mix of specific items and general categories. In all, we identified 75 discrete responses from across our sample, though not all were mentioned with the same frequency. Some were mentioned by nearly all respondents, whereas others were mentioned by only one or two. Table 2 lists those responses that recurred with relative frequency 4 . Because we were particularly interested in comparing stakeholders’ perceptions, we listed each stakeholder group separately. Responses from all five starter questions are collapsed in Table 2, because the questions were all 4 Relative frequency was subjectively defined, but in most cases, there was a clear threshold. For example, some terms elicited from one group might have two or three mentions, and then the next closest cluster had seven or eight. All inclusion decisions were made as a research team, and do not reflect the bias of a single person. designed to elicit different dimensions of the same concept. B e n ef ic ia ri es G o ve rn m en t A D R A Employment/Livelihood Opportunities X X X Good Infrastructure X X X Financial Security X X X Good Transportation X X X Support from the Government/NGOs X X X Peace and Security (Crime-Free) X X X Improved Living X X Well-Dressed People X X Organized/Clean Society with Working Systems X X Happiness/Good Attitude X Hard-Working Citizens X Community/Family Unity X Self-Control X Good Health Knowledge and Systems X X Good Governance X X Progress/Improvement X X Technology/Communication X Success X High-Quality People X Polite/Well-Mannered People X Intelligence/Knowledgeable People X Independent People X Table 2: Development is… Response extraction of beneficiary data yielded 39 distinct terms; 14 were recurring and appear in Table 2. Response extraction of government data yielded 31 distinct terms; again, 14 were recurring and appear in Table 2. Response extraction of ADRA employee data yielded 52 distinct terms; coincidentally, exactly 14 were also recurring and appear in Table 2. Looking at the table, one notices some overlap among the three stakeholder groups. But there is also some difference. Figure 1 is a Venn Diagram of how much overlap there is between stakeholder groups on Christian Relief, Development, and Advocacy 1(2), Winter 2020 Kiš, Cabus, Romo, and Welch, “We are poor, but we are developed” 99 the meaning of “development.” It is based upon the data presented in Table 2. Figure 1: Venn Diagram of Overlap Between Stakeholder Perceptions (A = Beneficiaries; B = Government; C = ADRA) From Figure 1, it becomes clear that slightly less than half (6 out of 14) of the perceptions of “development” are shared in common across all three stakeholder groups (ABC). If one compares any two stakeholder groups (AB + ABC, for example), slightly more than half of the perceptions are held in common. Thus, for each stakeholder group, slightly less than half of their perceptions are completely unique. Given that this was not a representative sample, however, one must be cautious about making overly extensive conclusions about any overlap in the general population’s perceptions on the meaning of the word “development.” Such confidence is better suited to the quantitative analysis which took place during the latter phase of this study (see “Quantitative Results” below). It should be noted that exclusion of certain terms that were hardly mentioned by a particular stakeholder group does not mean that nobody from that group mentioned it at all. Only that it was not a major recurring term. Thus, for example, one should not read Table 2 to say that nobody from the beneficiary group mentioned anything about good governance at all, only that it was not a major recurring term for beneficiaries, whereas it was for government workers and ADRA employees. Digging Deeper Having completed our primary task of distilling vast amounts of information into discrete terms, we looked again at the narrative descriptions our respondents provided during interviews and focus group discussions. Some illustrative quotes help place responses in context, and understand where respondents are coming from when they mention certain key terms. For example, when expounding upon her response that Employment (one of the terms listed in Table 2) is a key component of her definition of “development,” a 29-year-old female nutritionist officer from ADRA said that a developed person is someone “with a source of income, with a stable job, one who is independent and can provide for the people who are dependent on him without outside help or forces.” This also touches upon the themes of Financial Security and Independent People. Good Infrastructure, on the other hand, means “good highways, a nice [taxi] terminal, a baywalk, [public] comfort rooms, and a good marketplace.” according to a 27-year-old beneficiary housewife. “Many big houses, basketball courts, gymnasium, school buildings, and public comfort rooms,” added a 33-year-old beneficiary housewife. For one 33-year-old male ADRA employee, good infrastructure took a very specific shape: “Uniformity of houses.” The concept of support from the Government/NGOs, while being shared across all three stakeholder groups, appears to be perceived differently, even if it was mentioned in common. Many beneficiaries listed some form of external care as a sign of development, such as government projects to help poor people, ongoing support from international organizations, or remittances sent from relatives abroad. A 38-year-old beneficiary housewife said that, to her, development means getting “support from the government, many programs, and help from the president, especially during disasters.” Note that beneficiaries did not speak of outside support as a means to attain development, but rather as a sign of being a developed society, community, or individual. Developed people, in this conception, are those who are well-cared for on a long-term basis by benevolent external agencies. Contrary to Western conceptions of development bringing independence and self- sufficiency (see the following paragraph), many beneficiaries saw development as bringing ongoing dependence. Although this study did not set out to establish whether or not the Philippines is a clientelistic society, the belief that permanent dependence upon a wealthy benefactor is a marker of success is one hallmark of patron-client relationships, which is well- documented in the literature (Kiš 2014; Hutchcroft 1997; Auyero 2000; Sun, Zhu, and Wu 2014; Szwarcberg 2013). In contrast to the beneficiary view of social support, ADRA employees stressed that it was only to be a safety net for the poorest of the poor. A 35-year-old male ADRA employee emphasized that in a developed society, the government “supports those who are not well off,” but that such a society still “has poor people who haven’t taken full advantage of the opportunities afforded to them.” In other words, government support is a stop-gap measure, not a permanent fixture. Many Christian Relief, Development, and Advocacy 1(2), Winter 2020 Kiš, Cabus, Romo, and Welch, “We are poor, but we are developed” 100 ADRA employees mentioned the closely related theme of independence, stating that it is a necessary criterion for development. A 29-year-old female nutritionist officer said that “education is not a criterion because some are graduates but not developed because they are dependent.” Social support and independence are not contradictory in this view; those who are developed are independent and don’t need social support; but social support exists to take care of those who are not developed yet. Elaborating upon the idea of improved living, a 41- year-old beneficiary housewife said, “Development means there is hope to have a better life.” Well-dressed people, according to a 24-year-old beneficiary housewife, means, “clothing, jewelry, accessories, physical appearance or grooming, having an education, being fluent in conversation, and having a good attitude.” Happiness/good attitude was expressed by one 19-year-old female beneficiary student as “bundles of joy shared.” Not only did beneficiaries view dependence upon wealthy benefactors as a positive marker for development, they also favored dependence upon each other as a sign of good development. Many beneficiaries spoke of a developed society as being characterized by unity, sharing, cooperation, and caring within families and communities. A 52-year-old male farmer said that development means “being able to help others financially when needed, being able to easily lend money.” A 23-year-old female nutritionist officer described development as Good Health, saying that a developed person is “tall or the right height (or standard normal height), normal weight, has pimple-free skin, is knowledgeable about nutrition facts, has a complete set of teeth, and has clear vision.” Good governance was described by a 27-year-old male ADRA nutritionist as “commitment and involvement of community leaders to any program, even though it’s not profitable to them.” A 25-year-old female social worker said that in a developed society, you have “visible government officials.” When asked to elaborate, she said that meant that they were out in the community doing their jobs, rather than holing up in an office somewhere refusing to mingle with their constituents. And a 32-year-old male accountant said that in a developed society, people “trust local government officials; they can see where their taxes go, and there is less corruption, more peace, and more security.” One member of the government focus group alluded to Technology/Communication as “having nice gadgets,” which she elaborated to mean cell phones, computers, and tablets. This idea was reinforced by the other members of the focus group; throughout the discussion, reference to technology and communication came up repeatedly. The government group also talked about development being associated with high-quality people, or “good-looking people”, as one respondent put it. A developed person is one with “good looks and gains weight,” said another. Finally, when talking about Intelligence/Knowledgeable People, a 23-year-old female ADRA nutritionist officer said, “developed people are knowledgeable, living in reality.” Self-Perceptions of Development Given the wide range of definitions of “development,” it is perhaps not surprising that beneficiary answers to the final starter question – “Do you consider yourself to be a developed person? Why, or why not?” – were split almost equally three ways between “Yes,” “Partially,” and “No” (Yes = 7 respondents; Partially = 5; No = 6). Of those who answered “Yes” or “Partially,” they frequently cited having achieved a satisfactory level on one or more dimensions of development that had already been identified. For example, “We are poor, but we are developed,” declared a 62-year-old beneficiary housewife. “My children managed to find good work and to give an example to others; they respect us very well. We are poor in money, but rich in manners.” This response acknowledged her poverty in one sense – financial – while affirming her wealth in another – having good manners and respect. “Finding extra income is easy,” said a 35-year-old female small business owner. “Many people are coming and asking for help, I have a good source of income, I am able to pay my debts and shop for clothes, and I am able to provide for the needs of my children.” And a 36-year- old male fisherman stated that he is developed because “I am doing my best to improve so I can be able to buy the things I want; I am happy because I have a fishing boat.” Among ADRA employees, answers to the final question about personal development were mostly spread between “Yes” and “Partially” (Yes = 5 respondents; Partially = 4; No = 2). Reasons given for being developed included “Because I am independent,” “I have a good source of income,” and “I have (material possessions).” Those who were more tentative about their personal development status equivocated that “Development never stops; it’s continuous,” “I’m still learning and developing skills,” and “I’m still not developed enough.” Quantitative Methods The qualitative phase produced rich information that gave us insights into how development is perceived by the stakeholders under investigation. It also gave us Christian Relief, Development, and Advocacy 1(2), Winter 2020 Kiš, Cabus, Romo, and Welch, “We are poor, but we are developed” 101 a list of terms to be used to illustrate perceptions of “development.” But it cannot tell us the extent of any differences in perception, nor even confirm whether there is any difference at all. Since the respondents for the qualitative phase were not drawn using a probability sample, we cannot rule out that any apparent differences between groups are actually the result of sampling bias and not any real underlying variation. For that, we need a quantitative survey using an unbiased probability sample. That is the focus of phase two of this study. In order to test our hypotheses, we required a structured survey instrument that could be administered to each of the stakeholder groups, and that would detect similarities or differences of perception in the meaning of “development.” Moreover, we wanted a locally relevant instrument which referenced ideas and artifacts from everyday Filipino life. Finally, we wanted our survey prompts to be standardized. For these reasons, we decided upon a pictorial survey. With words (“Let me describe a scenario, and you rate how developed that image seems to be”), everyone would evaluate a different mental picture. With photographs, everyone would have the same image and would be evaluating the same survey prompt. Because a survey of this kind has never been done before in the same way and same place, we developed a pictorial survey of our own. In the qualitative phase, respondents had converted the images in their heads into words; for the quantitative phase, we converted those words back into images. First, we took all 75 responses – including those that did not recur – and grouped them into categories. As described in the beginning of the “Qualitative Results”, we noted that some of the responses were discrete items (such as “cell phones”) and others were categories (such as “technology”). For clarity, we categorized all responses; those that were already categories remained so, and those that were discrete items were classified into categories. This was conducted through pile sorting of the 75 responses written out on index cards and was a team effort. After we had pile sorted all responses, we ended up with 22 categories, some containing as few as one item, others containing as many as eight. This process guided us when developing the pictorial survey. We ultimately depicted discrete items and images in our survey, rather than abstract categories. But we selected the items in such a way that they represented the full spectrum of categories. This way, we could avoid having, for example, ten survey prompts about technology and none about health. After selecting the items for the pictorial survey, we staged a series of photographs with props and local volunteers (who gave their informed consent, as did all interview and survey participants) in order to depict the selected terms. For example, Figure 2 is the survey prompt for Category 6 (Technology/Communication), Item g. (Cell phones); Figure 3 is the survey prompt for Category 15 (Organized/Clean Society with Working Systems), Item a. (Garbage segregation). Figure 2: Survey Prompt 6g Figure 3: Survey Prompt 15a In some cases, we used several images to represent a category, such as 6.a. (Computers) and 6.g. (Cell phones), which both represented Category 6 (Technology/Communication). Notice also how the image in Figure 3 is local in origin; recycling and garbage bins around the world look very different from those in Leyte province. If we had simply used a word- based survey, differences in mental image might mean that respondents weren’t evaluating the same thing. In all, forty pictures were produced that depicted terms derived from the qualitative phase. We did not want the survey to be too short to detect significant differences in perceptions, nor too long to maintain respondents’ attention. When administering the survey, we showed each of the pictures (printed on large photo paper) to respondents one at a time (Figure 4), read the caption (such as “Cell phones” or “Garbage segregation”), and asked respondents, “On a scale of 1- 5, with 1 being very underdeveloped and 5 being very developed, how developed does this image appear to Christian Relief, Development, and Advocacy 1(2), Winter 2020 Kiš, Cabus, Romo, and Welch, “We are poor, but we are developed” 102 you? Why?” The “why” was a way of triangulating back to the first phase of the study to see if a new set of respondents picked up on the same themes that the qualitative respondents had identified. It was also a way of validating the survey itself; if respondents regularly identified other elements in the images unrelated to the caption that influenced their responses, we could know that the images were not good depictions of their underlying themes. In practice, this never took place. This confirmed the face validity of our survey instrument. In addition, five of the survey questions were reversed; that is, they depicted scenes that we expected to be rated as not developed (such as slums or street beggars). Inclusion of some reversed questions usually helps to increase the validity and reliability of a survey instrument because it measures the same concept from different angles. Figure 4: Survey Data Collection Finally, we shuffled the stack of pictures among respondents to eliminate ordering bias, which can occur when survey prompts are always given in the same order. People may be influenced to answer questions in specific ways if they are preceded by certain other questions; shuffling the order of question administration canceled out this ordering effect. Our initial Cronbach alpha reliability measurement was .896, which is exceptionally good. However, all five reversed items presented problems; four of them had negative item-total correlations, and one of them had a very low positive item-total correlation (.069). Our survey reliability analysis also indicated that the Cronbach alpha could be significantly improved by deleting these five items from the questionnaire. Having done so, we ended up with a Cronbach alpha of .924 and no remaining problematic items. What this means in practical terms is that survey respondents didn't consistently understand reversed questions. Far from improving the reliability of the survey, adding reversed items only muddled its comprehensibility. Thus, our final survey questionnaire contained 35 items that were reliably understood by respondents. As with the qualitative phase, quantitative surveys were administered to each of the three different stakeholder groups: Surveys of beneficiaries and government officials took place within the communities where ADRA was working, while surveys of ADRA employees took place at ADRA offices. The geographical separation of the primary locations of qualitative data collection (Bato) and quantitative data collection (Tacloban) minimized potential data contamination between the two groupings of respondents, as the two areas are approximately 150 km apart. The survey was administered to 249 ADRA project beneficiaries, fifteen local government officials, and thirteen ADRA employees. This represented the totality of ADRA employees not interviewed in the qualitative phase, the majority of local government officials liaising with ADRA projects in the area, and approximately 10% of the beneficiary population in the area (2,500 people were targeted by the ERL project in the vicinity of Tacloban). The sex ratio (female/male) for beneficiaries was 69.5/30.5, for ADRA employees was 35.7/64.3, and for local government officials was 73.3/26.7. As with the qualitative phase, professions of respondents varied from positions within ADRA to occupations within the communities, such as farmer, student, laborer, housekeeper, and small business owner. The average years of education for beneficiaries was 7.5 years, 14.5 for ADRA employees, and 9.1 for government officials. Quantitative Results After administering the surveys, we ran ANOVA in SPSS to compare mean responses across stakeholder groups on each of the 35 questions (see Appendix 1 for details). Out of 35 survey questions, there were 11 (31.4%) with significant differences between at least some of the groups under comparison, which represents nearly a third of the survey questions. Table 3 lists those variables with significant group differences, and also indicates which groups had significantly different means from the others. As can be seen from the table, question 6a was an anomaly; its test statistic indicated that there would be a significant difference in means between at least two groups, but its post-hoc test did not find any group differences. Nevertheless, for the remainder of the group differences, it is clear that the ADRA group always differed with at least one of the other groups, and nearly always differed from both groups. The beneficiary and government groups, however, never differed significantly from each other. ADRA was clearly the standout group. Christian Relief, Development, and Advocacy 1(2), Winter 2020 Kiš, Cabus, Romo, and Welch, “We are poor, but we are developed” 103 Q u e st io n C ap ti o n G ro u p D if fe re n ce s5 2a Getting Support from Different Organizations 3a(2) Having Public Comfort Rooms 3e Having a Nice House 3h Having a Public Market 5c Owning a Business 6a Having Computer Access ??? 8e Having Basic Furniture 12a Studying Well for School 14a(2) Having a Hospital Nearby 20a Having Bus Service 22a Owning Land Table 3: Significant Differences Between Groups Recall that the hypotheses under investigation were as follows: H0: There is no significant difference between stakeholder groups on perceptions of the meaning of “development.” H1: There is a significant difference between stakeholder groups on perceptions of the meaning of “development.” Thus, Table 3 is not a measure of agreement, but of difference. The remaining 24 out of 35 questions that do not appear on Table 3 were those for which no significant differences appeared. The ANOVA results 5 B = Beneficiaries; A = ADRA; G = Government. If groups are listed with a line between them, that means there was a significant mean difference between those groups. for each of the 35 questions could be viewed as testing 35 distinct sub-hypotheses; for some questions, the null hypothesis was rejected, whereas for others, the null hypothesis failed to be rejected. In order to make a determination on the overall hypothesis, we needed to look at the 35 questions collectively. For this, we ran MANOVA (see Appendix 1 for details). The MANOVA test was highly significant. This means that when considered collectively, we can answer our overall research question: There is a significant difference between stakeholder groups on perceptions of the meaning of “development.” We reject the null hypothesis. As we saw with ANOVA, about a third of the survey questions were responsible for this overall significant result. Thus, stating that there is a significant difference between stakeholder groups on perceptions of the meaning of “development” does not mean that their perceptions differed in all cases, or that they always disagreed. It simply means that their perceptions differed often enough to be statistically significant. Discussion and Conclusions The qualitative phase provided useful terms for quantitative survey elaboration, as well as demonstrating the nature of differences in perceptions on key terms associated with the meaning of “development.” The quantitative phase confirmed that significant differences in perception about the meaning of “development” do exist between stakeholder groups, and those differences were always manifested between the ADRA group and the others. Two particularly noteworthy findings are that beneficiary respondents focused on personal attributes (rather than structural ones) when defining “development,” and beneficiaries perceived “development” as meaning greater dependence on a benefactor (be it ADRA or the government) rather than greater independence. The reasons for these noteworthy findings could be varied. Although we did not measure the wider population during this study, the default assumption is that ADRA beneficiaries do not significantly differ from the general population in their values and perceptions, meaning that these ideas and attitudes are widespread. Indeed, the fact that the beneficiary group never differed from the government group (who were not ADRA beneficiaries) suggests that beneficiary perceptions are not solely a function of being recipients of ADRA’s aid. Further evidence of ADRA’s significantly different perception of what “development” means is the fact (already noted) that two-thirds of beneficiary respondents answered “yes” or “partially” when asked if they themselves were developed persons. One can Christian Relief, Development, and Advocacy 1(2), Winter 2020 Kiš, Cabus, Romo, and Welch, “We are poor, but we are developed” 104 presume that if ADRA is a development agency and it has selected those individuals as beneficiaries, then ADRA probably does not share the view that they are developed. What could account for those differences in perception? Some possible explanations: 1. ADRA employees’ level of education is much higher than the average population. This is by virtue of the fact that ADRA jobs require minimum levels of education that exceed the average in the Philippines. In school, ADRA employees may have been exposed to more new ideas. 2. Once a person is employed by a development agency, much effort is typically expended to train them to see the world through the organization’s lens. ADRA employees may have simply learned to see things differently than their compatriots through explicit training. 3. Finally, the ADRA group reported significantly more exposure to developed contexts than either of the other two groups. 6 Perhaps exposure to developed countries alters one’s perception of what is possible, and what a developing country should aspire to. Regardless of how one explains the differences in perception between ADRA and the other groups, an additional important question centers around what those differences mean for ADRA’s programs. How can ADRA’s means and ends reflect that the organization takes the perspectives of other stakeholders seriously? Without being privy to the internal workings of ADRA Philippines, we (the authors) do not presume to say whether or not ADRA Philippines already does this, but we do suggest that development organizations working anywhere in the world should at least consider the possibility that their perceptions of “development” may differ from those of their fellow stakeholders. Responses that take those differences seriously could include confirming stakeholder perceptions via a study (much like ours) and then crafting means and ends in such a way that they satisfy multiple stakeholder priorities. As a practical example, if ADRA Philippines aims to diversify livelihoods for the sake of disaster resilience whereas beneficiaries desire to diversify benefactors that they can depend upon, then both ends can potentially be met by not only training beneficiaries in new professions (e.g. training some farmers how to run small businesses) but also by plugging them into existing 6 One demographic question asked respondents to indicate if they had ever been to what they would consider a developed country. Although a few beneficiaries and government employees responded in the affirmative, the vast majority had not, whereas about half of ADRA employees had. professional networks (potential benefactors) which support those professions (e.g. a local small business association). In other words, taking beneficiary perspectives seriously doesn’t have to mean completely revamping programs and abandoning development agency goals. It can mean simply expanding program objectives to include goals that matter to beneficiaries as well. Beyond explaining differences and considering how to take those differences seriously in programmatic terms, a third dimension is what those differences mean for evaluation. Supposing one studies whether any differences in perception exist among stakeholders on the meaning of “development,” and supposing an organization adjusts its programs to satisfy multiple stakeholders, what will be the measures of success? Will the organization focus merely on those aspects of development that interest its donors, or will the organization also hold itself accountable to beneficiary expectations? Much has been written about how development agencies run the risk of feeling beholden only to their donors, as that is the source of continued financing and existence (see Easterly 2007). Faith-based NGOs such as ADRA are not immune from these pressures, but we suggest that this is an area where faith- based NGOs can take the lead in promoting and demonstrating an ethical treatment of fellow stakeholders than is greater than the norm. This is one way that a faith-based orientation can add value – integrating beneficiary community concerns into all phases of the development cycle (including evaluation) instead of simply going through the motions as a sort of tokenism just to please donors. This threefold approach to addressing differences in perception when they are discovered – explaining them, integrating them into development programs, and allowing them to inform and expand measures of success – can be applied at multiple levels. The most immediate level, of course, is the organization that participated in this investigation, ADRA Philippines. Next, the worldwide ADRA network may consider how to apply this approach elsewhere. Finally, other agencies (particularly faith-based ones) may contemplate whether this is an appropriate model for them to follow as well. This study was noteworthy not only for its findings, but also for what it didn’t find. Specifically, it is curious that spirituality did not appear significantly in respondent answers about what “development” means to them. ADRA is a faith-based NGO and the Philippines is an overwhelmingly Christian country (the only Christian-majority nation in all of Asia). While Christian Relief, Development, and Advocacy 1(2), Winter 2020 Kiš, Cabus, Romo, and Welch, “We are poor, but we are developed” 105 spirituality did register in respondents’ answers, it did not register significantly, as one would expect. Table 2 included only those items that had recurring mentions when voluntarily freelisted by respondents. And Table 3 reported only those items for which there were statistically significant differences between groups. During freelisting, at least one respondent did mention that developed people are those who pray more often and are more faithful in going to church; but this was a rare mention, and therefore didn’t make the cut for Table 2. Nevertheless, the research team thought it might still be worthwhile to include at least one spiritual item on the quantitative survey, so we included an image of a woman praying as one of our survey prompts for respondents to evaluate on a scale of 1-5 (very undeveloped to very developed). This spiritual item did not make it onto Table 3 either as being one of the items that differs between groups, but this does not mean that it wasn’t rated as highly developed. In fact, all three groups rated “prayer” as a 4 or above on average, meaning that they considered it an aspect that was either “developed” or “very developed.” Since Table 3 was looking for significant differences, and all stakeholders rated spirituality similarly, it didn’t appear on Table 3, but this does not mean that it wasn’t considered important. It is not unprecedented that an item that was so rarely mentioned during freelisting scored so highly when presented as a fixed survey option. People do not always immediately think of things that are important to them when asked to freelist. For example, if one were to ask respondents who the most important people are in their lives, few respondents might list their boss, but if one presented respondents with fixed response options which included “boss,” more respondents might rate their boss as highly important. Another explanation is also plausible: given ADRA’s non- (religiously)-discriminatory stance in programming and hiring and its non-proselytizing orientation, respondents may have been initially reflecting what they thought ADRA wanted to hear. In fact, given that many NGOs operating in the Philippines are non-sectarian, perhaps the default assumption on the part of beneficiaries is that their benefactor is secular in nature and not interested in hearing about (or addressing) respondents’ spiritual welfare. Only when the door is opened to the spiritual dimension by the benefactor do beneficiaries feel comfortable to go there themselves. This has implications for faith-based NGOs. If they desire to address beneficiaries’ spiritual needs as well their physical ones, they must be explicit about their openness to address spirituality. If, on the other hand, faith-based NGOs wish to focus primarily on the physical, leaving the spiritual aspect in the realm of motivations, then beneficiaries may perceive them as secular. Clear communication in this area may help in better aligning understandings of “development.” In conclusion, definitions and understandings of what it means to be developed differ significantly between stakeholder groups, specifically between NGOs and the others (beneficiaries and government liaisons). Recognizing and addressing these differences is a stepping-stone to more ethical humanitarian efforts. References Arndt, H. W. 1981. “Economic Development: A Semantic History.” Economic Development and Cultural Change, 29(3):457—466. Auyero, J. 2000. “The Logic of Clientelism in Argentina: An Ethnographic Account.” Latin American Research Review 35(3):55—81. Bell, Simon, and Stephen Morse. 2011. “Sustainable Development Indicators: The Tyranny of Methodology Revisited.” Consilience, 6(1):222— 239. Blowfield, Michael. 2005. “Corporate Social Responsibility: Reinventing the Meaning of Development?” International Affairs (Royal Institute of International Affairs 1944-), 81(3):515—524. Bräutigam, Deborah. 2011. “Aid ‘With Chinese Characteristics’: Chinese Foreign Aid and Development Finance Meet the OECD-DAC Aid Regime.” Journal of International Development, 23(5):752—764. Campbell, Lisa M., and Arja Vainio-Mattila. 2003. “Participatory Development and Community- Based Conservation: Opportunities Missed for Lessons Learned?” Human Ecology, 31(3):417— 437. Carroll-Bell, Sam. 2015. “Development Alternatives in Timor-Leste: Recasting Modes of Local Engagement.” Bijdragen tot de Taal-, Land- en Volkenkunde, 171(2/3):312—345. Cavalcanti, Joabe G. 2007. “Development versus Enjoyment of Life: A Post-Development Critique of the Developmentalist Worldview.” Development in Practice, 17(1):85—92. Chandy, Laurence, and Homi Kharas. 2011. “Why Can't We All Just Get Along? The Practical Limits to International Development Cooperation.” Journal of International Development, 23(5):739—751. Cornwall, Andrea, and Karen Brock. 2005. “What Do Buzzwords Do for Development Policy? A Critical Look at 'Participation', 'Empowerment' and 'Poverty Reduction'.” Third World Quarterly, 26(7):1043—1060. Christian Relief, Development, and Advocacy 1(2), Winter 2020 Kiš, Cabus, Romo, and Welch, “We are poor, but we are developed” 106 Creswell, John W., and Vicki L. Plano Clark. 2006. Designing and Conducting Mixed Methods Research. Thousand Oaks, CA: Sage. Diallo, Amadou, and Denis Thuillier. 2004. “The Success Dimensions of International Development Projects: The Perceptions of African Project Coordinators.” International Journal of Project Management, 22:19—31. Easterly, William. 2007. The White Man's Burden: Why the West's Efforts to Aid the Rest Have Done So Much Ill and So Little Good. New York: Penguin Books. Eyben, Rosalind, and Rebecca Napier-Moore. 2009. “Choosing Words with Care? Shifting meanings of women's empowerment in international development.” Third World Quarterly, 30(2):285—300. Gallie, W. B. 1956. “Art as an Essentially Contested Concept.” The Philosophical Quarterly (1950-), 6(23):97—114. Heryanto, Ariel, and Nancy Lutz. 1988. “The Development of ‘Development’.” Indonesia, 46:1—24. Holmberg, J., and R. Sandbrook. 1992. “Sustainable Development: What Is to Be Done?” In Making Development Sustainable: Redefining Institutions, Policy, and Economics. Edited by Johan Holmberg. Washington: Island Press. 19—38. Horn, Philipp, and Jean Grugel. 2018. “The SDGs in Middle-Income Countries: Setting or Serving Domestic Development Agendas? Evidence from Ecuador.” World Development, 109:73—84. Hutchcroft, P. D. 1997. “The Politics of Privilege: Assessing the Impact of Rents, Corruption, and Clientelism on Third World Development.” Political Studies 45(3):639—658. Ingham, Barbara. 1993. “The Meaning of Development: Interactions Between ‘New’ and ‘Old’ Ideas.” World Development, 21(11):1803— 1821. Kiš, Adam D. 2014. “Non-Political Clientelism: A Lesson in Confusion for Sustainability.” OIDA International Journal of Sustainable Development 7(7):75—80. Lewis, David, and David Mosse. 2006. “Theoretical Approaches to Brokerage and Translation in Development.” In Development Brokers and Translators: The Ethnography of Aid and Agencies. Edited by David Lewis and David Mosse, West Hartford, CT: Kumarian Press. 1— 26. Lipi, Rahima K. 2016. “Dignity and Empowerment: An Exploration of the Microcredit Experiences of Women in Rural Bangladesh.” Journal of International Women's Studies, 18(1):230—259. Marcus, Richard R. 2001. “Seeing the Forest for the Trees: Integrated Conservation and Development Projects and Local Perceptions of Conservation in Madagascar.” Human Ecology, 29(4):381—397. Offutt, Stephen, and Amy Reynolds. 2019. “Christian Ideas of Development: Understanding the Current Theories, Networks, and Priorities of Accord Organizations.” Christian Relief, Development, and Advocacy 1(1):1—14. Parkinson, Sarah. 2009. “Power and Perceptions in Participatory Monitoring and Evaluation.” Evaluation and Program Planning, 32:229—237. Quinn, Claire H., Meg Huby, Hilda Kiwasila, and Jon C. Lovett. 2003. “Local Perceptions of Risk to Livelihood in Semi-Arid Tanzania.” Journal of Environmental Management, 68:111—119. Scheld, Patrick. 2018. “Who Controls Haiti's Destiny? An Examination of Haiti's Underdevelopment, Endless Poverty, and the Role Played by NGOs.” Master’s thesis: The University of Vermont. Seers, Dudley. 2010 [1979]. “The Meaning of Development.” In Development Theory: Four Critical Studies, edited by David Lehmann. New York: Routledge. Sen, Amartya. 1999. Development as Freedom. New York: Oxford University Press. Sun, X., J. Zhu, and Y. Wu. 2014. “Organizational Clientelism: An Analysis of Private Entrepreneurs in Chinese Local Legislatures.” Journal of East Asian Studies 14:1—29. Szwarcberg, M. 2013. “The Microfoundations of Political Clientelism: Lessons from the Argentine Case.” Latin American Research Review 48(2):32—54. Tipple, Graham, and Suzanne Speak. 2005. “Definitions of Homelessness in Developing Countries.” Habitat International, 29:337—352. Zimmerman, Felix, and Kimberly Smith. 2011. “More Actors, More Money, More Ideas for International Development Co-operation.” Journal of International Development, 23(5):722—738. Adam D. Kiš is an Associate Professor of Anthropology at Burman University in Canada. Dr. Kiš has also lived in Benin, Guinea, Sao Tome and Principe, Madagascar, the Philippines, and the United States. He worked for ADRA for six years. Author email: adamkis@gmail.com Jovy Conde Cabus is an independent consultant who has worked with a variety of organizations over the years, including ADRA Philippines. She is currently a mailto:adamkis@gmail.com Christian Relief, Development, and Advocacy 1(2), Winter 2020 Kiš, Cabus, Romo, and Welch, “We are poor, but we are developed” 107 tutor at RareJob, Inc., an organization that teaches the English language to Japanese clients. Author email: cabusjovy26@gmail.com Tzaddi Romo is an independent research consultant in the Philippines. He holds degrees in Secondary Education and Theology. He has worked in the Marshall Islands and has served on the editorial staff of the Research Journal of Mountain View College. Author email: romotzaddi7@gmail.com Cheyanne Welch is a recent graduate of Burman University’s International Studies program. She is a freelance journalist who has lived in Egypt, the Marshall Islands, and the United States. She has worked for A Better World Canada as an intern. Author email: cheyannewelch@burmanu.ca mailto:cabusjovy26@gmail.com mailto:romotzaddi7@gmail.com mailto:cheyannewelch@burmanu.ca Christian Relief, Development, and Advocacy 1(2), Winter 2020 Kiš, Cabus, Romo, and Welch, “We are poor, but we are developed” 108 Appendix ANOVA Details Because sample sizes were not equal among the three groups, we had to be careful about which procedures we used. Levene’s test told us whether each question violated the assumption of homogeneity of variance. If it did not, then we used the usual F-test to any significant differences among groups, and we used Hochberg’s GT2 post-hoc test – which is recommended for different group sample sizes when equal variances are assumed – to identify which groups were different from each other. If Levene’s test did indicate that a particular question had violated the assumption of homogeneity of variance, we then used Welch’s F to identify significant differences among groups, and we used the Games-Howell post-hoc test to identify which groups were different from each other. Both Welch’s F and the Games-Howell post-hoc test are robust to violations of the assumption of equal variances. Ultimately, however, all the ANOVA tests measured the same thing. MANOVA Details Rather than a single F-statistic as in ANOVA, MANOVA has four tests: Pillai’s Trace, Wilks’ Lambda, Hotelling’s Trace, and Roy’s Largest Root. These all test a single hypothesis which has multiple components. Significant debate continues over which test statistic is most appropriate to apply in various situations, but these are irrelevant to our particular case, as all four test statistics were highly statistically significant (p<.000).