An operational research framework is a tool to conceptualize the sequence of process steps in integrated natural resources management. It provides concrete process steps, which can be followed implementing integrated natural resources management at local resource users’ level (Hagmann et al. 2002,19). Although social scientists prefer a more loosely structured approach to data collection, i.e. a framework that emerges in the course of the study, this rather applies for experienced researchers with plenty of time (Miles and Huberman 1994,17). In reality and being new to qualitative studies a tighter design is a wise course providing clarity and focus (Miles and Huberman 1994,17). However, there is no general notion how tight or loose a research framework needs to be (Miles and Huberman 1994,17). Miles and Huberman (1994,17) notice that this depends on how much is already known about the phenomenon to be studied, where to look for it and how to gather the information. They continue, that the task of arriving at a set of analytical categories may be done deductively from the start or inductively in a step-by-step process. The development of such an operational framework to suit integrated natural resources management in practice has had various inputs. Harrington et al. (2001,3) suggest the use of ecosystem analysis, focusing on genetic and environmental productivity, sustainability assessment, the definition of problem-cause relationships and how these affect system productivity and sustainability. Van Noordwijk et al. (2001,11) suggest “integrated models” to illustrate process steps implementing integrated natural resources management in tropical forest margins focusing on policy analysis. However, major contributions came from Campbell et al. (2001) who developed a raw model of a framework to help indicate the numerous entry points for intervention and performance assessment of natural resource systems (Figure 15). Their framework requires the researcher to clearly state his or her research objective, define spatial and temporal scales, and identify particular intervention possibilities (Campbell et al. 2001). They further suggest focusing on the key relationships among system components and the constraints impacting on them (Campbell et al. 2001).

Figure 15. Components of integrated natural resources management.

Source: adapted and modified from Campbell et al. (2001); Swift et al. (1994)

Figure 15 reflects different factor areas and the multiple dimensions of stakeholders and scales, which are a main characteristic of integrated natural resources management systems. The arrows illustrate the multiple scales of interaction and response (Campbell et al. 2001). Moreover, the framework takes into account that natural resource management systems involve multiple stakeholders with multiple perceptions and objectives and therefore multiple management strategies (Campbell et al. 2001). It is these characteristics the researcher needs to consider while paying particular attention to identify integrated natural resources management aims and also find trade-offs that meet the often-contrasting stakeholder interests (Campbell et al. 2001). In this context Allen (1997,633) emphasizes that ‘cooperation’ is far more realistic than ‘consensus’ and that with help of participatory processes, it may be possible to “develop win-win strategies”.

The illustrated components of integrated natural resources management give a rough outlook of the natural resource system and its resource users. The following illustration of an integrated natural resources management learning cycle helps the researcher to understand how integrated natural resources management research is implemented and which activities the researcher must cope with to accomplish the sharing of results and the evaluation of the ongoing integrated natural resources management research process best described as ‘learning by doing’. The figure of the learning cycle embodies the essence of adaptive management, participatory action research and the conceptual framework (Figure 16). It is the direct answer to the natural resource system and its management, which is in a constant state of flux, hence demands to be adequately responsive to the situation. This implies that research does not start with a precise question and methodology but that the questions and the appropriate methodology arise from the study (Dick 1993,11). Dick (1993,11) continues “let the data decide” meaning that as the process of integrated natural resources management unfolds along the learning cycle, the information available at each step is used to determine the next step. Good process facilitation is essential to implement the integrated natural resources management learning cycle and move along its steps while developing appropriate solutions together with resource users (Hagmann et al. 2002; Sayer and Campbell 2001). Sayer and Campbell (2001), who developed the learning cycle in integrated natural resources management research through enhancing Hagmann’s et al. (2002,3) illustration of a framework with the sequence of process steps and the action research cycle, define the following steps: 1) exploratory phase or subsystem identification; 2) reflection and negotiation phase; 3) implementation or action phase; 4) evaluation phase; and 5) updating, i.e. readjustment and adaptation (cf. CGIAR 2003,9).

Figure 16. Implementation of the integrated natural resources management learning cycle.

Source: Sayer and Campbell (2001)

The implementation of integrated natural resources management evolves across spatial and temporal scales. As a result integrated natural resources management research does not involve one learning cycle but depends on a number of interlinked learning cycles (Sayer and Campbell 2001). However, depending on the phenomenon some may not complete a single learning cycle, whereas others will have been through many cycles within the project timeframe (Sayer and Campbell 2001).

Both, the framework of integrated natural resources management components and the integrated natural resources management learning cycle, are concepts to develop a specific operational framework that integrates across disciplines, across scales, across stakeholders, and across components (Lal, Lim-Applegate and Soccimarro 2001,2), which will vary depending on the research project. As mentioned in the previous chapter the researcher needs to consider which component to add into his operational framework carefully weighing the additional costs and the potential benefits against each other (cf. Campbell and Hagmann 2003,4). However, complexity is not boundless and “three to five key variables often drive any particular system” (Holling et al. 2000, cited in Sayer and Campbell 2001), thus “allowing system complexity to be reduced after an initial more inclusive phase” (Gunderson and Holling 2002, cited in Campbell and Hagmann 2003,5).

The operational framework used in this study and the process steps associated with the appropriate research methods are described in greater detail in this section. The presented frameworks are modifications of natural resource management approach and offer a comprehensive overview of the relevant elements in a coastal fishery environment. In addition they exemplify the multitude of factors that need to be considered and the roles the researcher needs to adopt when conducting integrated natural resource management. Furthermore, integrated natural resource management of coastal and inland fisheries has been relatively neglected so far (cf. Payne 2000,1).

In their handbook on the rapid appraisal of fisheries management systems Pido et al. (1996,8) present a framework for data collection and analysis consisting of contextual variables and their attributes of a coastal community’s local-level fisheries management system. The framework sets out to identify the biophysical, technical, market, socioeconomic, institutional, organizational, and exogenous variables affecting the fishery system (Pido et al. 1996,12). The contextual variables and their attributes in the framework from Pido et al. (1996,9-16) are similar to the integrated natural resources management components presented in the framework from Campbell et al. (2001), although Pido’s et al. work misses the multi-sectoral dimension and the multiple scale of stakeholder interactions. In exchange they offer detailed characteristics of the fisheries management system and operational considerations, e.g. required researchers and field data collection techniques. Pido et al. (1996,12) collect attributes associated with the contextual variables following guide questions about:

• Who is involved in a situation?

• What are their stakes and resources?

• What type of action can they take?

• What is the cost of those actions?

• What information is available to them?

• How much control do individuals have?

• How are individuals and actions linked to one another and to outcomes?

• What outcomes can be expected in relation to what action?

• How are rewards and punishments allocated to particular combinations of actions and outcomes?

Pido et al. (1996,12) identify 33 attributes in six sets of contextual variables and present adequate techniques for field data collection and personnel requirements for the composition of a multidisciplinary research team (Table 2).

[Source: adapted and modified from PIDO et al. (1996,9)]

Table 2. Contextual variables and their attributes, field data collection techniques and the required researchers.

The biophysical and technical attributes (Group I) provide information over the natural resource system, its productivity, exploitation level and the employed fishing techniques. The market attributes (Group II) focus on the market structure, prices and their fluctuation. The characteristics of stakeholders (Group III) refer to the socioeconomic status, livelihood strategies, resource conflicts and stakeholders perceptions that affect their incentive to cooperate. Group IV relates to institutional and organizational arrangements inside the communities, i.e. arrangements concerning access to and use of natural resources for example. Group V relates to institutional and organizational arrangements outside the communities, i.e. at national, provincial, district or municipal levels, that regulate or restrict the use of natural resources that are responsible for decision-making, policy development and implementation. Pido et al. (1996,13) point out that there “may be nested, multiple layers of organizations, formal or informal, at different political and administrative levels”. Finally, exogenous factors (Group VI) relate to sudden shocks that affect the management system, i.e. weather, political or cultural changes.

To further emphasize the role of the researcher in integrated natural resources management particularly the ‘integrated’ aspect of sectoral and disciplinary considerations Silvestre (1996,6) offers a schematic representation of an expanded multisectoral and multidisciplinary framework for fisheries management, which he uses to demonstrate the key constraints in integrated fisheries management efforts (Figure 17). The following framework originated from ICLARM research activities in San Miguel Bay, Philippines, from 1979 to 1982 and from 1992 to 1994. The framework (Figure 17) highlights the human and natural dimension in coastal fisheries. Silvestre (1996,12) stresses that, “successful management requires good communication and agreement among three groups - the managers, the resource users or stakeholders, and the scientists”. Furthermore he emphasizes that the key constraints pertain to sectoral and disciplinary “blinders”, i.e. the human perception seeing parts rather than the relevant whole being conditioned by sectoral interests and disciplinary background.

Figure 17. Schematic representation of an expanded framework for fisheries management.

Source: Silvestre (1996,6)

Influenced by the previous frameworks for fisheries management systems I developed the following operational framework with the sequence of process steps (Figure 18). Beginning with a preparation and planning phase including desktop studies and the organization of the fieldwork I quickly moved into the field to conduct a situation analysis (Figure 18). After learning more about the resource system and its multi-stakeholders the next step included the development of shared visions to establish platforms for further procedure. By means of participatory methods and facilitation and mediation of interests participants were encouraged to envision sustainable fishery management options. Through sharing of results the collected data was verified concerning the validity of the studied resource system. The self-evaluation of past experience was an important step to re-adjust the research process, including research design, applied methods, shared problems, and the fishery management options collected so far. It is important to understand that the research cycle and the sequence of process steps were not followed stringent but instead that process steps were repeatedly passed through, i.e. process steps recurred in repeating sequences. The outcomes are various: (i) first and foremost awareness building among participants laying the basis for shared decision-making to successfully introduce fishery management strategies; (ii) second of all and not subject of this study planning and implementing integrated natural resources management; and (iii) last but not least publications and the present thesis (Figure 18).

Figure 18. Operational research framework with sequence of process steps.

Source: adapted and modified from Hagmann et al. (2000,27)

The following section provides an overview of the applied methods and techniques during the six-month investigation period in Vietnam. All methods and techniques were extracted from the three-volume sourcebook “Participatory Methods in Community-based Coastal Resource Management” published by the International Institute for Rural Reconstruction (IIRR 1998). At first, qualitative methods such as observation, identification of key informants, semi-structured interviews, focus-group discussions, resource mapping, and institutional analysis were chosen to establish partnership between the various stakeholders and learn more about the coastal fishery sector, rural livelihoods, and existing problems. In addition more quantitative methods were selected including problem ranking, Venn diagrams, and questionnaires. The combination of qualitative and quantitative methods is an important and recognized technique that leads to an improved quality of information (Marsland et al. 2001,14).

The selection of methods was based on personal experience with participatory methods, the achievement of research objectives, and the suitability of creating a learning-cycle for awareness building and self-reflection, and the feasibility of methods according to the utilization of financial and human resources in a given time frame.

The selected tools and techniques were modified to the circumstances at hand, i.e. cultural and societal characteristics. In addition methods were adapted according to scarce resources particularly personnel and time. For example, the questionnaire was not handed out to participants as planned but filled out by a research assistant in a process of reading the questions out loud to the respondents and filling in their answers, because high illiteracy and limited reading skills threatened the quality and return of the questionnaires.

Observation is probably the most simple and direct empirical method to gain insight into a community and the processes within it. By observing what actually happens in a community it is possible to understand how it operates. Observation took place anywhere the subject was found, e.g. during fishing activities, at landing sites, at processing or marketing facilities, government agencies, and in households. The observations were carried out taking notes immediately afterwards or if customs prevented, e.g. at formal meetings especially when government officials attended, memorized observations were written down on the very same day. Observation not only took place in the beginning of the field study but during the entire project especially when conducting participatory methods and techniques to enhance learning, e.g. while drawing resource maps or facilitating group discussions, where after a brief introduction I withdrew from the exercise or the discussion and observed the unfolding events. The relatively unsystematic gathering of information through observation provides the researcher with preliminary data necessary for developing more refined research methods like semi-structured interviews, questionnaires or matrix-ranking techniques (McGoodwin 2001). Furthermore, observation established the basis for developing relationships with the stakeholders and thus for interviews and follow-up visits. By structuring the observations i.e. choosing surroundings and social settings, I considered central for the coastal fisheries resource system, focusing on the identification of the key drivers, the key response variables, and the key intervention points of the problem being addressed, namely the degradation of coastal resources, and through writing down observations immediately afterwards, the method of observation became systematized and thus scientific (cf. McGoodwin 2001). Observation is more than simply gathering information in the field and passively recording what people are doing and saying (McGoodwin 2001). The researcher needs to combine the observed events and behaviors with additional information from further questions or literature according to his personal frame of reference (McGoodwin 2001). In addition the researcher needs to be aware that his presence affects the social interactions, e.g. my presence as observer in particular as a foreigner focused the overall attention and distracted people from their daily activities but as I became more and more accepted by the communities and stakeholders, my influence on the observed people and events declined. In order to avoid biased observations the researcher also needs to be aware, that he or she selects and notes down that which precisely supports his or her research hypotheses. To prevent selective observation I used my research framework as guideline and formulated research hypotheses much later in the research process. The strength of the observation is certainly to observe social interactions, which would not occur in an interview. For example, I saw how fishermen using illegal fishing techniques were separated from the social life of the village, discriminated, and how they formed their own groups. The strongest limitation of using observation as a method is its demand of resources, i.e. time and effort. Transportation, agreements, cancellations, and the time spend for the actual observation take up a large amount of time. In addition a fair amount of time was spent for noting down the observed.

Observation is a powerful tool of data collection and retrieving first hand information to adjust the further research process to the situation at hand and complement qualitative research methods. However, its efficiency depends on the researcher’s abilities to interact with people and to deal with new and changing situations regarding culture, society, politics, and geography.

The identification of key informants is an important step to gather relevant information and utilize scarce resources in the research process. The IIRR (Vol. 2 1998,61) defines key informants as “purposely selected community members who are able to provide information on a particular research topic based on their knowledge, skills or experience”. The purpose of using key informants is to obtain accurate, relevant, and detailed information about the community or from an individual community member without talking to everybody (IIRR Vol. 2 1998,62). This is also one of the first steps before applying other participatory methods such as semi-structured interviews, resource mapping, or Venn diagrams. The authors assume that “chairpersons and members of a fishers’ organization for example can give detailed information about issues confronting fishers in the locality (e.g. marketing, illegal fishing) using their experience as basis”, but although this may apply in most cases, there is a danger of selecting biased information and drawing a distorted picture. For example, I detected that the information from interviewed chairpersons was tinted according to the view of the Communist Party and in sharp contrast to the information obtained from other group members because only party cadres and functionaries acted as chairpersons or village leaders. In my field study the identification of key informants was carried out moving down the hierarchic ladder, i.e. visiting the government authorities, members of the Politburos, chairmen of the People’s Committees, village leaders and finally individual community households and members. The reason for this was that I had to apply for a permit to visit each government district that gave me access rights to the different villages and letters of invitation. After following the formal procedures and building first contacts other key informants were identified within the communities usually with help of the village leader. According to the suggested approach in the IIRR (Vol. 2 1998,64) sourcebook a meeting was arranged with help of the village leader to explain the purpose of the field study. Preceding that the village leader was asked to identify community members that hold key positions in the fishery, aquaculture, and agricultural sector. An attempt to include women in this procedure failed due to sexual inequality (gender issues). As a result women were approached directly at home or during fish processing and marketing activities at local markets. The village leader also helped in providing telephone numbers and addresses of the identified key informants. In the next step the key informants were selected according to their relevance to provide information for the research objective. Another method of identifying key informants was to visit the people whose names I heard repeatedly, e.g. during group discussions, semi-structured interviews, and through visits of People’s Committee meetings.

The biggest limitation in the process of identifying key informants is to detect those informants that give misleading or biased information. For example I noticed that in many incidences the opinion of key informants was in line with the Communist Party ideology. By crosschecking the information by asking many informants the same questions this problem was avoided. Another limitation is that some persons dominate participatory processes which lead to biased assertions or prevent people from expressing their opinion, e.g. in some villages the village leader dominated these processes. In such cases facilitation skills were applied to help participants express their opinions, but not to offend local values, and key informants were visited privately.

The identification of key informants is a quick and powerful tool to get to know many individuals in a certain area and context. The knowledge that may be gained from key informants should not be underestimated especially about more complex issues that may be not appropriate for group discussions (DFID 2000). However, by means of triangulation, asking a number of people the same questions, the researcher needs to crosscheck information to avoid bias.

The IIRR sourcebook (Vol. 2 1998,78) defines semi-structured interviews as a conversation with a purpose that differs from a structured interview with a specific set of questions. In a semi-structured interview there is only a set of guide questions or discussion points, e.g. fishery degradation, illegal fishing, or enforcement and the interview evolves in response to the interview situation and the participant’s assertions (cf. Deffner 2004,3).

The purpose of the semi-structured interview is to generate information by means of leaving the development of an interview to the interviewed individual and his or her personal experience.

Selecting possible interview partners follows the same approach as in the identification of key informants, as described in the previous section. However, I frequently selected interview partners randomly on the street, in cafes and markets, or during boat and net repairs, where it was obvious that the approached persons will be able to provide relevant information on the research subject. Through this procedure I could also make sure that the participants felt at ease conducting the interview in their familiar surroundings.

The questions should be phrased, so that informants are required to tell a story. The IIRR sourcebook (Vol. 2 1998,79) gives four examples:

1. Descriptive questions – to prompt informants to describe their daily activities. Example: “Could you describe what happened after the fish traps were constructed?”

2. Structural questions – to find out how informants understand their situation and acquire knowledge.
   Example: “Has the use of cyanide affected your fish catch? If so, how?”

3. Contrast questions – to encourage informants to discuss situations and provide an opportunity for comparison.
   Example: “How would you compare the fish catch today with that of 10 years ago?”

4. Probing questions – to allow informants to analyze and reflect on the underlying causes of a particular situation.
   Example: “Why?” and “How?”

After a short introduction I asked the informants, if he or she had some time to answer a few questions. I explained the purpose of the interview and gave a brief overview of my research. The semi-structured interview was started with general questions about the informant’s family and household and then moved to more specific questions encouraging him or her to become more descriptive. In order to deepen the conversation questions were asked in different ways especially by use of probing questions. Usually the interview was written up simultaneously. This procedure was only possible, because interviews were carried out with a translator and the time during translation was utilized this way. Most interviews lasted between one and two and a half hours. At a later stage in the field study I used the semi-structured interviews to feed information from the research project back to the participants. Again, just like in the other applied participatory methods combined with the learning cycle inherent in integrated natural resource management, this led to deep self-reflection and further insights of the complex natural resource system and its users.

The strength of semi-structured interviews is its responsiveness to the individual and the situation at hand. Besides gathering information it can generate perceptions and emotions (IIRR Vol. 2 1998,81). After some experience semi-structured interviews are easy to apply and require no additional preparation time or appointments. For example, I carried out semi-structured interviews between fixed appointments or whenever the situation allowed.

A limitation of semi-structured interviews is certainly that responses may be influenced by biases (IIRR Vol. 2 1998,81), e.g. informants interviewed in cafes most likely responded differently due to the surrounding people listening. On the other hand interviewing people privately does not guaranty, that the answers given are not what they think you expect to hear. Therefore applied interview technique puts a high demand on the interviewer and his communication and mediation skills incorporating the accumulated information into the interview process and establishing a form of triangulation.

Group discussions, also referred to as focus group discussions, are discussions with a selected group of community members (key informants or others chosen for their relevance to the objective of the study) following a guideline designed to generate discussion on a particular topic (IIRR Vol. 2 1998,45). The purpose of group discussions is to gather information on livelihood practices, decision-making structures, issues in fishery or agriculture and other information (IIRR Vol. 2 1998,45). In addition information previously collected during group discussions may be verified or detail added.

I used group discussions to create a learning cycle, i.e. information from previous visits and different communities was shared among the participating group in order to enhance people’s self-reflection and raise awareness. On other occasions group discussions were used to reach a consensus between conflicting groups or when discussing contentious issues.

Preparation for a group discussion started with inviting target participants, e.g. drag net fishermen or shrimp farmers. These were usually selected together with the village leader paying attention that group members represented various income categories or community organizations. As mentioned before bias is a problem and too homogenous groups should be avoided. After briefly introducing my translator and myself I presented the purpose and time frame of my research. The introduction varied considerably since I held several group discussions in each village and community members started to recognize and become acquainted with me making the introduction redundant. The actual group discussion was started with any non-controversial topic raised by the participants to warm up. To initiate discussion open-ended questions were asked.

For example:

What could be done to improve fishery production?

Who should enforce regulations and how?

How do you envision your future?

Questions were also phrased to encourage self-reflection and a personal responsibility for the problems the community members expressed, through mirroring back to people “the consequences of their present perceptions and behaviour and possible solutions in the long run” (Hagmann et al. 2000,9). During the discussion I offered cookies, because bringing along small presents is a custom in Vietnam and I also discovered that this habit also acted as an icebreaker. The entire discussion was handled flexibly according to how it developed. In order not to lose focus my research framework (goals and objectives) acted as a guideline. However, due to the complexity of the coastal fishery sector and following an action research approach the course of a group discussion was very much left to the participating community members, i.e. the discussed topics usually in one way or the other referred to fisheries and I only participated in the discussion as much as necessary, e.g. providing suggestions when the matter was of no importance or acting as a mediator when discussions went overboard. During group discussions participants sometimes left to take care of their duties or sometimes new faces would join a discussion.

The strength of group discussion is certainly the generation of information, which is verified and clarified at the same time. Even discussing apparently simple topics, as for example, which fish drag net fishermen catch, lead to deeper insights learning that the neighboring village did not use any drag nets compared to the first village where all of the fishermen engaged in mostly illegal drag net fishing. Group discussions are not only important to understand livelihoods but they promote learning and lead to awareness building if following a participatory learning approach. For example, towards the end of my field study participants frequently suggested possible approaches how to manage natural resources, a phenomenon nonexistent in the beginning of my study. The limitation of applying group discussions in integrated natural resource management is its personal requirements, i.e. group discussion demands a great deal of facilitative skills, a clear understanding of the research framework, and specific knowledge of the addressed disciplines (cf. Hagmann et al. 2000,10).

Resource mapping is a method of visualizing the occurrence, distribution, access, conflict over, and use of resources within a specific area (IIRR, Vol. 2 1998,222). The resource map provides a topographic overview and generates further information during the making. Resource maps are helpful for internal discussions, for planning and monitoring purposes, and support the researcher in the conduct of in-depth assessment (IIRR, Vol. 2 1998,223).

The purpose of drawing a resource map is to allow community members to identify and map critical locations such as areas known for illegal fishing, pollution, siltation, mangrove deforestation, etc. The map also helps the researcher to learn more about the particular view of a community’s physical boundaries and spatial distribution of issues. Last but not least, drawing a resource map serves as an ice-breaker and helps to develop a relationship with the community and for this reason was conducted early in the field study but only after a few introduction visits. The IIRR sourcebook (Vol. 2 1998,224) endorses this adding “resource distribution, use and access may be considered a sensitive issue by the community” and highlights that “knowledge of the social structure of a community is a prerequisite for the facilitator”.

The following materials were used:

craft paper (1x2m)

pencils in assorted colors

felt tip pens in assorted colors

markers

The use of a topographic map as suggested in the IIRR sourcebook was cancelled, because it drew too much attention away from the actual map-drawing exercise and discouraged people to draw freehanded.

The mapping exercise was planned together with the village leader who invited the same community members as for group discussion. According to the suggestions of the sourcebook I explained that the map will stay in the community and I will only make a small copy. In my opinion this step is vital to show the participants that they are not being taken advantage of, but rather through a consultative way of participation the collegial role between researcher and clients is strengthened and approved (cf. Probst and Hagmann 2003,11; Susman and Evered 1978,597). Next the purpose and scope of the exercise was described to the group and then the craft paper spread out on the floor. Then the group was asked to select individuals, who they thought, are knowledgeable about certain resources or sectors, e.g. drag net fishing, gill net fishing, aquaculture, etc. The selected individuals were told to draw the outline of the lagoon first, locate their village, and add landmarks, e.g. roads, mountains, islands, rivers, etc. together with their names. After finishing the first step and receiving confirmation by the group the pencil marking was fixed with permanent marker. In the next step the participants were asked to locate fishing areas, spawning grounds, corals, mangroves, conflict areas, areas with aquaculture, specific species areas, and specific gear fishing areas on the map and draw them in any order they chose. In addition I asked participants to use different symbols and colors and generate a corresponding legend. During the resource mapping the rest of the group was encouraged to comment and correct the map drawers. This process varied considerably from active participation to dominance by the village leader. After completing the map and verifying it by the entire group the pencil marking was fixed with felt tip pens.

Documenting the resource mapping exercise is an essential part of the output and plays an important role. For this reason I did the introduction to the exercise and then my translator took over the supervision to allow for thorough process observation and documentation. The composition of the map and its features reflect the perception and relationship between the participants and the natural resources (IIRR Vol. 2 1998,227). For example, the most important fishery resource will appear first on the map, just as issues on resources like illegal fishing will appear exaggerated in size and color, an observation I can confirm (IIRR Vol. 2 1998,227).

The strength of resource mapping is that it visualizes the distribution and use of natural resources in a certain area. The exercise itself helps to improve the relationship with the community (cf. IIRR Vol. 2 1998,229). Besides visualization the process also generates information and provides insights that may not have been detected otherwise. In addition it triggers self-reflection and awareness building of participants and thus participatory learning. Another major advantage is that I became familiar with the area, e.g. fishing sites and conflict areas in a particularly short time. However, a limitation to resource mapping was the amount of time it demanded especially in view of the scarce personnel resources.

Institutional analysis is the identification of various resource users, stakeholders and organizations respectively their institutional arrangements in a community (IIRR Vol. 3 1998,118).

The purpose is to identify existing legislation, policies and regulations that regulate use and access of coastal fishery resources and in particular to determine the decision-making processes of who decides and implements institutions governing coastal resource management at various government levels, i.e. village, municipal, district, province, national (IIRR Vol. 3 1998,119). Evaluating the existing involvement of local resource user groups in the management of coastal fishery resources is vital to determine key intervention points for the involvement of resource users in future integrated natural resource management. This is a delicate process where the researcher needs to be aware of the existing structures and that the incentive to improve resource management and encourage resource user participation in this process threatens existing power structures hence is unpopular.

In my field study institutional analysis took place throughout the entire study period. During the first step secondary data was collected from documents, reports, publication, national policy papers and fishery laws both national and provincial documents. In a second step informants from communities, organizations, and government offices were added to collect primary data using semi-structured interviews, group discussions, resource mapping, and Venn diagrams. The IIRR (Vol. 3 1998,121) sourcebook and Pido’s et al. (1996,46) handbook for rapid appraisal of fisheries management systems suggest the use of guide questions at this point and offer a wide range. However, I pursued an investigative action research approach constantly adjusting and improvising as the events unfolded and according to the circumstances and whatever interview partners suggested. Therefore the guide questions only served as an orientation in the back of my head and data about institutional arrangements was collected whenever and however the situation allowed. The collected data was verified according to compliance, complementariness and conflict to ensure accuracy and to fill in gaps. Yet community members often lacked information and government officials were secretive and retained information.

This method provides information about formal and informal institutions governing coastal fishery resource management in Nha Phu Lagoon.

The strength of institutional analysis is that it provides the researcher with knowledge that he may not consider to collect otherwise and through sharing this with the communities and other stakeholders in the research process, resource users may deepen their understanding about resource management and government officials may learn about practices of resource users which otherwise are not available. Especially when developing management strategies together with participants, how to manage coastal fishery resources more sustainable, integrating bio-physical and socio-economic elements with institutional information proved vital (Strehlow and Peters 2004,5).

The limitation to institutional analysis is that it requires a skilled facilitator due to the complexity of the matter (IIRR Vol. 3 1998,129).

Problem ranking, also referred to as matrix ranking, is a tool to identify and rank problems according to a set of criteria within a community (IIRR Vol. 2 1998,21).

The purpose of problem ranking is to allow community members to prioritize their problems and thus focus their attention and resources to the most important problems (IIRR Vol. 2 1998,21). In my study the problems that negatively affect the degradation of fishery resources were identified during previous visits by means of focus-group discussions, resource mapping, and semi-structured interviews. Moreover, the problem ranking exercise was part of a questionnaire that I conducted in all 12 villages covering six households in each village. Its format therefore differs widely from the problem ranking exercise as described by the IIRR sourcebook using Meta-Plan technique. Participants were encouraged to name further problems to those mentioned in the questionnaire. Next the participants were asked to compare each problem and rank it from one to ten according to its relevance on the degradation of fishery resources. The problem with the highest number is considered the number one problem. If several problems have the same number, they are considered to be of equal importance.

The strength of problem ranking is that it provides a systematic way of prioritizing problems in a community (IIRR Vol. 2 1998,26). Problem ranking is also a powerful tool for awareness building in a community, because it reflects back problems gathered during previous visits and thus is part of implementing the integrated natural resource management learning cycle.

The limitation of problem ranking is that it is time consuming and the ranking procedure requires thorough explanation. In addition the researcher should be aware that problem ranking may raise expectations in a community and is best followed by planning and implementing a course of action (IIRR Vol. 2 1998,26).

Venn diagrams are a tool for illustrating relationships between different groups and organizations within a community (IIRR Vol. 2 1998,116; Pido et al. 1996,33). They are particular useful to identify potential conflicts between stakeholders (Pido et al. 1996,33) but also potential key intervention points to introduce more sustainable resource management practices. For example, Venn diagrams can identify different decision-making patterns concerning resource management and which stakeholder group is responsible for making decisions about a given resource (Schoonmaker Freudenberger 1994).

The purpose of Venn diagrams is to identify groups and stakeholders and the relationships between them and the community (IIRR Vol. 2 1998,117). Furthermore, Venn diagrams identify the influence of organizations on each other and a particular subject in the community, e.g. in this study the economic development of fishery households

I chose the ‘economic development of family households’ as the center of the contemplation, because people in different communities perceived the influence of stakeholder groups and organizations on the community itself variedly, according to income, fishing gear, education, location of their village in the lagoon, etc. only the influence on their own financial situation was perceived to be somewhat equal.

• Circles – symbolize the organizations and stakeholders in the coastal fishery sector.

• Size – is determined through ranking and resembles the influence of the identified organizations or stakeholders on the local coastal fishery.

• Position – means the distance of the organizations or stakeholders to the economic development of fishery households and represents the relationship between them and the fishery households. In addition overlapping circles or circles in close proximity represent business ties and relationships between stakeholders, organizations and/or fishery households.

The following materials were used:

craft paper (1x2m)

colored paper cut into three sizes of circles

markers

glue

The Venn diagram was conducted in three representative villages of the 12 villages situated around Nha Phu Lagoon. Due to lack of personnel resources and time the course of preparing a Venn diagram was modified pursuing the following process steps:

1. A meeting was arranged and previous selected key informants, between four and eight participants, invited. Particular emphasis was paid to invite people from various professions that bear reference to the coastal fishery sector.

2. The objective and use of the Venn diagram was explained, including the notice that the original diagram will stay in the village and only a small copy made.

3. In a next step the participants were encouraged to complete a list of previously selected stakeholders related to the coastal fishery sector by either adding or crossing out stakeholders.

4. The list of stakeholders was ranked according to the ranking technique in the previous section. The ranking resembles the influence of the stakeholders on the local coastal fishery ranging from one to ten.

5. In the meantime the three sizes of circles were labeled with the stakeholders in relation to their rank, i.e. the larger the circle the more influence the stakeholder group represented by that circle has on the issue (see Figure 19), namely:
   1-3 small
   4-7 medium
   8-10 large.

6. The craft paper was placed on a table, the participants provided with the prepared circles and the subject economic development of family households drawn as a large circle in the middle of the craft paper (Figure 19).

7. Then the participants were asked to place the circles according to:
   • their influence/relationship on the economic development of the family households and
   • their influence/relationship on each other through its relative distance from the middle or through overlapping each other (Figure 19).

8. After allowing enough time to discuss the positioning of the symbols and letting the participants review their outputs, the circles were glued to their final position.

9. The finished Venn diagram was then object of a thorough discussion about the role and influence of stakeholder groups on the livelihoods of fishery households and the entire coastal fishery sector (Figure 19).

Figure 19. Sample output of a Venn diagram.

The strength of Venn diagrams is that it summarizes complex information in a visual manner, identifies informal groups and other stakeholders which have not been detected in the previous research process, and that it reveals the local perception on relationships and the level of influences of stakeholder groups and organizations (IIRR Vol. 2 1998,123). Besides producing detailed information about stakeholders and their influence the process provides the participants with a better understanding about the most powerful and influential key stakeholders affecting their livelihoods. I found the Venn diagram technique precious for learning more about power relations inside the coastal fishery sector without provoking people to take a stand and retrieving biased information but instead producing many lively discussions.

A weakness of Venn diagrams is that it consumes a lot of time and requires educated participants who are able to transfer their knowledge to the abstract concept of circles, sizes, and positions. Due to the perceived high rate of illiteracy among fishermen and thus the encounter of difficulties conducting the Venn diagram, the selection of participants orientated itself along the educational level instead of selecting a representative group of the community. In addition conducting Venn diagrams also puts a high demand on the researcher, i.e. thorough preparation, the capability to explain matters, and the facilitation of the long and multi-faceted process of guiding people but not dominating them. Although the exercise marked the end of the field study in Vietnam and is good for fostering the integrated natural resource management learning cycle to enhance reflection, create awareness and prompt discussions, I experienced that comprehensive participatory research is a prerequisite before applying this tool.

Questionnaires are a method to collect data using prearranged questions in a specific order or in other words a written interview. A questionnaire is designed in line with the research project, i.e. the objective of the study, the overall circumstances, cultural and social factors.

The purpose of a questionnaire is to collect information that provides a genuine reflection of the attitudes and beliefs of people. Thereby the questionnaire gives the researcher greater confidentiality, because the interaction between the respondent and the researcher is limited, so that the questions can be answered more thoroughly.

The questionnaire was carried out in each of the 12 villages around Nha Phu Lagoon, with respectively six households, totaling 72 households. The households were selected using a stratified approach. The objective was to cover sales persons, middlemen and representatives in agriculture, aquaculture and fishing particularly of the most common gear-fishing categories in each village.

Respondents were explained the purpose and use of the questionnaire and assured that their answers will be handled confidentially. Due to widespread illiteracy the questionnaire was read to the respondents and filled in by the research assistant. Hence questions concerning the length, design, distribution and return (response rate), and attractiveness of the questionnaire were not an issue. However, the language used in the questionnaire is kept simple and questions easy to read.

The question structure in the questionnaire survey comprises a combination of open-ended and closed questions with the majority being closed questions to assess demographic and fishery data. Within the closed question range there are a number of response options from the simple Yes/No choice to multiple-choice questions including an ‘other’ exhaustive category offering a wide range of options for selection. Moreover, order questions were included in the questionnaire to perform a problem ranking (see also problem ranking in the previous section).

The strength of questionnaires is that they save time by allowing a large number of individuals to be questioned. Questionnaires also encourage objectivity, because confidentiality can be guaranteed and all questions are presented in an identical order. In addition the bias was somewhat controlled, because all 72 questionnaires were filed out by one person.

The weakness of questionnaires is that they are solely extractors of information with little or no use to the respondent. This is because questionnaires are inflexible and one short question follows the other giving the respondent little time to think and reflect his answers and the researcher no possibility to interrogate the respondent to clarify answers.

In order to reach the objective of creating a learning cycle and override some of the weaknesses mentioned above, I experienced the inclusion of the problem ranking exercise into the questionnaire as a major improvement.

Empirical data were collected using the ‘choice of methods’ mentioned above. In addition desk-based literature reviews were carried out prior to the field study and throughout the entire research process. The literature review mostly relied on web-based materials such as scientific journals, project reports, guidance sheets, best practice guidelines, articles, and abstracts of books to capture the global state of knowledge. On the other hand, more use was made of printed material, i.e. books, journals, project reports, and official documents. The latter were acts, decrees, and code of conducts published by the Vietnamese government particularly from the local government and related to coastal fishery management and aquaculture.

The research design concerning the selection of theories and paradigms, philosophy of science, methodological issues and varying qualitative approaches was largely influenced by websites committed to such areas. Inputs from sites such as http://nrm-changelinks.net (action research), www.iatp.org/AEAM (adaptive management), http://forum.objectivismonline.net (epistemology) or www.qualitative-research.net (qualitative social research), only to name a few, proved vital to encourage critical and reflective thinking.

In general data, collected during group discussions, participant observation, or resource mapping exercises, was entered into a journal directly afterwards in form of a summarized transcript. On other occasions I abandoned this procedure. This occurred in particular in my role as facilitator or mediator. In addition a lot of information was obtained casually after meetings and exercises, during lunch invitations, at local coffee shops or during long motorbike rides. This information was written down in field notes and then converted into write-ups at a later moment. Through this procedure some of the missing information were added back, because reviewing of field notes often lead to reminiscence of information (cf. Miles and Hubermann 1994,51). Complementarily over 230 photographs were taken to augment field notes and to provide a more comprehensive understanding of the coastal communities. By means of writing summarized transcripts or write-ups the amount of data was reduced considerably in advance compared to commonly used elaborate transcripts. This approach ensured that a fine balance was achieved between the amount of information generated and the number of participants engaged in the research process. Miles and Huberman (1994,10-11) point out that data reduction occurs continuously throughout the research process even before data collection. They see data reduction as part of analysis and suggest, that scientist be conscious about their decisions and about which conceptual framework, which data collection procedures, and which data chunks to choose.

Approximately 242 respondents were involved in this research including the 72 respondents from the questionnaires. The respondents represent 21 stakeholder groups from four administrative levels: 1) village level, 2) local government level (municipal government), 3) local government level (district government), and 4) national government level (province government). At the village level representatives are fishermen, women, shrimp and lobster aquaculturists, middlemen, village leaders, Women’s-, Farmer’s-, and Fishermen’s Unions, and the Fishery Resource Protection Group. At the municipal government level the participants come from the People’s Committees and Councils and the Offshore Fishing Boat Support Group. At the district government level, representatives come only from People’s Committees and Councils. At the national government level these include representatives from the Ministry of Fisheries, Fisheries Resource Protection Department, the Fishery Promotion Centre, the Marine Border Police, and the Institute for Oceanography, the Research Institute for Aquaculture No. 3, the University of Fisheries, and banks (Table 3). However, respondents from Women’s-, Farmer’s-, and Fishermen’s Unions and few other stakeholder groups are not listed independently, because these participants joined group discussions, resource mapping and Venn diagram exercises. Therefore, their number lies within the accumulated number of fishermen (Table 3).

No

Stakeholder group

Number

Percentage (%)

Village Level:

01

Fishermen

153

02

Women

31

03

Shrimp farmers

15

04

Lobster farmers

8

05

Middlemen

2

06

Village leaders (Village authority)

12

07

Women’s-, Farmer’s-, and Fishermen’s Union

08

Fisheries Resource Protection Group

1

Local Government:

Municipal Level:

09

People’s Committee

1

10

People’s Council

11

Offshore Fishing Boat Support Group

District Level:

12

People’s Committee

3

13

People’s Council

1

National Government:

Province Level:

14

Ministry of Fisheries

1

15

Fisheries Resource Protection Department

1

16

Fishery Promotion Centre

1

17

Research Institute for Aquaculture No. 3

3

18

Institute for Oceanography

3

19

Marine Border Patrol

2

20

University

2

21

Banks

2

Total

242

Table 3. Distribution of respondents.

By means of triangulation the validity and reliability of the collected data was assured. Triangulation is the application and combination of research methodologies to study the same phenomenon. Besides the triangulation of methods there are other forms of triangulation including triangulation of observers, theories, empirical materials and data (cf. Fielding and Fielding 1989,25). The idea is to increase the accuracy of information through working with, preferably independent or partly independent, multiple information sources and determining similarities or differences between the data sources (Dick 1993,11). Dick (1993,11-12) distinguishes similarities between triangulation and dialectic and gives examples how to create dialectic using:

• different informants, e.g. asking fishermen and illegal fishermen the same questions;

• different research settings, e.g. visiting all twelve villages around Nha Phu Lagoon;

• the same informant responding to different questions, which address the same topic from somewhat different directions, e.g. what fishing technique do you use? How big is your boat and how big is the engine? Two questions to determine if the fisherman engages in illegal fishing, while keeping in mind that the motorization of boats corresponds to the applied fishing techniques;

• information collected at different times, e.g. visiting villages and fishermen at weekly intervals;

• different researchers or;

• different methods, e.g. asking questions that cover the same topic in semi-structured interviews as well as in resource mapping exercises.

Therefore triangulation of empirical data and qualitative and quantitative methods leads to more profound knowledge about the natural resource system and its users (cf. Flick 1992,195). However, triangulation is no panacea or guarantee of validity, choosing the appropriate methods and avoiding bias, i.e. selecting field data that fits the personal perception of the research matter, remain the responsibility of the researcher (cf. Fielding and Fielding 1989,32). Hammersley and Atkinson (1983,199) concretize that one should not assume that “the aggregation of data from different sources will unproblematically add up to produce a more complete picture”. In addition it is impractical and unwise to pursue all feasible means of data collection and analysis and use as many as possible (Fielding and Fielding 1989,34). Instead it is important to choose “at least one method of data collection that describes and interprets the context in which the interaction occurs and one that is designed primarily to illuminate the process of interaction itself” (Fielding and Fielding 1989,34). For example I used focus-group discussions and observations to learn about the problems that degraded fishery resources and combined them with Venn diagrams that revealed the various stakeholder groups and their interactions inside the coastal fishery sector.

Integrated natural resource management borrows and combines paradigms theories and methods from various research approaches to address natural resource systems. I continued this practice applying a mixed-methodology design to data analysis mixing qualitative and quantitative paradigms at many methodological steps (cf. Creswell 1994,178;). This procedure combines the advantages of both the qualitative and the quantitative paradigms, e.g. working back and forth between inductive and deductive models of thinking to better understand the concept being explored (Creswell 1994,178).

Greene, Caracelli and Graham (1989, cited in Creswell 1998,175) highlight five major purposes that address mixed-method evaluation design:

1. Triangulation – to test the consistency of findings obtained through different instruments (cf. previous chapter).

2. Complementarity – to enhance and clarify facets of a phenomenon with the use of other method-type results.

3. Development – to shape subsequent methods or steps in the research process, where the first method is used to help inform the second method.

4. Initiation – to stimulate new research through challenging of results and open up new perspectives.

5. Expansion – to provide breadth and scope to a study exploring specific features of each method and bringing their different inquiry components together in a side-by-side fashion (cf. Greene 1997,3).

A research strategy integrating different methods is likely to produce results that are potentially more useful and relevant in terms of quality and scope in contrast to research concepts following only one method (cf. Greene 1997,2-3; Mayring 2001,31; Burzan 2004,25). Although authors like Seipel and Rieker (2003) point out that the integration of qualitative and quantitative approaches cannot be done insouciant and demands a great deal of methodological expertise from the researcher, their biggest concern lies in the progression of qualitative and quantitative methods and the generally small sample size of qualitative research that does not allow statistical generalization (Burzan 2004,24). Mayring (2001,7) argues that the sample size is no clear demarcation criteria, since there are quantitative individual case studies. He continues, that up until now no explicit definitions exist that distinguish between qualitative and quantitative analysis. However, this distinction cannot be done by listing methodological differences but must start at a level regarding epistemology and philosophy of science (Rost 2003,41). Creswell (1994,175) expresses this notion through asking whether paradigms should or must be linked with research methods.

There are three stances in the debate about mixing paradigms: (1) the purists argue that paradigms and methods should not be mixed; (2) the pragmatists see paradigms as useful conceptual constructions but of little value for the research practice, and that researchers should make the most efficient use of methods to maximize contextual responsiveness; and (3) the dialectic stance in which paradigms are seen as valuable frameworks for the inquiry process and the tension that exists between two conflicting or interacting paradigms is viewed as potentially prolific generating a “more complete, more insightful, even more revisioned or transformed evaluative understanding” (Greene 1997,2). However, the same author pleads for “a balance between philosophy and methodology, between paradigms and practice” (Greene 1997,2).

Mayring (2001,5) emphasizes applying strategies for analysis according to their strengths and searching for strategies combining qualitative and quantitative methods. In the case of qualitative oriented research projects and adding quantitative analysis the research results gain generalization, transparency, and the research as such becomes comprehensive und revisable (Mayring 2001,31). Furthermore systemizing and generalizing provide arguments against doing research of mediocre quality (Mayring 2001,31; cf. Grønhaug and Olson 1999,6).

The research design aims to evaluate the data through qualitative and quantitative methods according to the model of triangulation (Figure 20). Thereby, the idea is not to determine which approach for analysis produces the better results, rather the results mutually support each other and the intersecting individual findings represent the overall results (Mayring 2001,25). The focus does not lie on the establishment of truth but to stepwise expand the cognition through comparing different approaches (Mayring 2001,25).

Figure 20. Integrating qualitative and quantitative analysis in the model of triangulation.

Source: Mayring (2001,21)

In order to understand the selection of adequate methods of analysis I emphasize that:

1. In farming systems research explorative research designs generally are evaluated and depicted in a quantitative and descriptive manner (cf. Bitsch 2000,9).

2. Integrated natural resources management offers only limited concepts how to systematically evaluate qualitative research data.

Due to the explorative research design applied in integrated natural resources management and participatory action research, I selected the grounded theory approach according to Glaser and Strauss (1979) as concept of inquiry, because it sees itself as strategy to discover new phenomenon and develop new theories in a methodological controlled manner while demanding an open-minded approach to the study object (cf. Deffner 2004,9). However, since the scope of the study was to implement integrated natural resources management in a coastal fishery environment and not to generate theory, only the first of three steps in grounded theory was applied (for further discussion on this decision, see Glaser 1992; Strauss 1994; Strauss and Corbin 1996; Böhm 1994). In an interview with Legewie and Schervier-Legewie (2004,59) Strauss refers to this step as “theoretical sampling” a synonym for starting early in the research process with evaluation of interviews and writing of memos to produce theories and hypotheses that determine the selection of the next interview partner or comparison group. Mayring (1999,83-84) calls this process “stop and memo” and refers to this term as writing memos of central aspects discovered during fieldwork and classically during participant observation. For the further evaluation of the collected data I choose qualitative content analysis according to Mayring (1999, 91), which is specifically well suited to systematically analyze large quantities of textual data. In addition, its coding procedure is well described.

Qualitative content analysis is one social science approach developed some 20 years ago to systematically analyze texts (Mayring 2000,1). This method ensures a methodological controlled and progressive analysis of data such as write-ups, summarized transcripts, or written memos (Mayring 2000,5). Thereby the data material is viewed line by line and phrases, sentences, or paragraphs are coded into corresponding categories. Categories are developed inductively from within the textual material

This coding procedure is similar to the “open coding” phase in grounded theory (cf. Deffner 2004,10-11).

Figure 21. Model of inductive category development.

Source: adapted from Mayring (2000,11)

The outcome is a set of categories (codes). Specific words, sentences, or text passages are dedicated to each category. Further analysis either interprets the category scheme regarding research questions and underlying theories, or coded text passages can be evaluated quantitative (Mayring 1999,94). Quotations from the textual material serve as clarification of a category and respectively exemplify explanations. The quotations are cited using an ID provided by the software Atlas.ti^®, for example (14:1). Thereby the first number refers to the document, i.e. the fourteenth (14:) document, and the second number refers to the chronological order in which the quotation was created, i.e. the 1^st (:1) quotation that was created in this document. For further reference the quotation IDs and their original terms, as well as their assigned codes are listed in the appendix.

The possibility of a statistical evaluation of the individual coded categories was discarded. The frequency of entries and how often a category is assigned does not allow any assertions over the reviewed context. Free quotations were used to exemplify findings and theories.

Qualitative analysis software represents a computerized alternative to doing analysis by hand. The main advantage of using qualitative analysis software is to structure the large volumes of qualitative data (O’Flaherty and Whalley 2002,4). However, Glaser (1998,185-186, cited in Fernández 2004,91) warns about falling into the “technological traps” of data analysis software, so that he is concerned about the unnecessary restrictions they create, the possibility that they inhibit the development of the researcher’s own skills, and the time-consuming learning curves they impose. In my work I use the software Atlas.ti^® which is conceptually linked to the qualitative data analysis approach of grounded theory (Muhr 1994,318). I used Atlas.ti^® to perform open coding of the textual data, select quotations for later reference, generate family codes, i.e. grouping of subcategories into categories, and write memos. Similar to Fernández’ (2004,91) experience I quickly learned to appreciate the use of Atlas.ti^® in my study as on the one hand it provided a fast way of coding and retrieving data records using the code manager without having to shuffle through large quantities of paper and on the other hand it did not impose a significant learning curve to work with the software. At first all data is transferred to a ‘hermeneutic unit’. The hermeneutic unit contains all qualitative analysis steps, i.e. generated codes, quotations and memos (Figure 22).

Figure 22. Qualitative analysis software (Atlas.ti^®) showing a group discussion write-up with coding stripes to the right of the screen, the code manager, a highlighted quotation, and two memos (ME).

Although Atlas.ti^® offers automatic coding tools, the entire material was coded manually following qualitative content analysis as described earlier. Thereby a text passage is marked with the mouse pointer. Subsequently one or more codes are assigned to the citation, whereas either a selection is made from the code list in the code manager or a new code is assigned (see Figure 22). Next the coded categories were grouped into code families with help of the family manager. The entire process was documented by writing memos in the memo manager. In my opinion one of the major advantages of working with qualitative data analysis software such as Atlas.ti^® is that underlying text passages of codes can be easily retrieved and viewed for comparison.

Quantitative data analysis was applied through the statistical evaluation of the 72 questionnaires that were carried out in the final phase of the fieldwork. The outcomes are used to give proof of or reject the results from the qualitative data analysis thereby adding to generalizing results and theories in this study. As measure of the most common central tendency the arithmetic average or mean was applied. Adding the values of all observations and then dividing that sum by the number of observations calculates the mean.