 |
| - Research
Statement |
 |
 |
 |
 |
 |
|
Attributed to: Julie
Newton 1. What is the process research? 1. What is the process
research?
The aim of process research is to understand some of the key processes and relationships that different individuals and households engage in as they seek to achieve wellbeing outcomes. It uses a range of qualitative research methodologies with a sub-sample of different individuals and households together with a re-analysis of the existing WeD data to discern the types of processes that are important in formulating wellbeing goals and strategies. The research into the processes in which different individuals and households engage in during the pursuit of wellbeing provides insights into the relationships between wellbeing outcomes and structures.  2. Conceptual rationale for Processes The process research is informed by two of WeD’s theoretical foundations: the Resource Profiles Approach and the Regime Approach
3. How it contributes to WeD research Process research is underpinned by a focus on the person (in the development of this element of the research we used the term Person-Centred-Process-Research). Processes refer to people acting in relationships with others and within specific social contexts. Any action (or interaction) has material, social, moral and symbolic dimensions, and the challenge is to move beyond discipline-based biases that examine one dimension with little awareness of the significance of the other dimensions. Different people (men, women, young, old, for example) are located differently in social structures and as such engage differently in social processes in their pursuit of security and wellbeing. Some processes are engaged in through individual actions and others are more collective in character. The research questions that drive the process research are:
4. Description The process research in WeD has involved two different approaches employing two different sets of research methods: the ‘thematic’ and the ‘core case’ approaches. The ‘thematic’ approach involves the exploration of a series of prominent and contemporary development issues that have been identified as important to the pursuit of wellbeing in the communities studied. These themes were identified by a combination of insights from the prior research in the communities as well as by taking cognisance of development debates and discourses within each country. A sample of different individuals and households are then interviewed on their process experiences in relation to these themes. Bangladesh, Peru and Thailand used the ‘thematic’ approach to explore the following themes
Both approaches share similar methods using a range of qualitative methods. They also rely on the development of ethnographic case studies on the processes and relationships people engage with to achieve wellbeing outcomes. Specific data collection methods include participant-observation, semi-structured interviews, life histories, diaries, and focus group discussions. Where possible, case studies are being selected to include individuals and households covered by other research components (especially RANQ) so as to permit cross-analysis. 5. How it was developed The choice of particular methods for generating process data was dependent on the overall research strategy adopted by each country research team. As was the case with the community profiles, a key priority was to ensure that each country team had sufficient flexibility in the choice of methods to explore processes in ways appropriate to the topic and in that context. BANGLADESH The Bangladesh team used the ‘thematic’ approach to explore the following themes:
In terms of livelihoods, the Bangladesh team decided to focus on the use of income and expenditure at the level of the household. The primary source of data was the income and expenditure survey. After the first two rounds of the survey, they realised that the data highlighted a need to look briefly at processes of borrowing and lending. The decision to focus on politics and community institutions was driven by the team’s initial observations of the sites. There is a rich variety of institutions and some of these have a significant bearing on the ability of people to pursue their wellbeing. In one site a particular institutional form known as mastaans (local gangs and mafia) was dominant and this resonated with important developments at the national level. This justified the decision to explore the emergence of this and associated institutions in more depth. Research into marriage and family relations built directly on the early phases of the quality of life research in which family relations were identified as being central to people’s quality of life. Family relations cover a range of key issues such as identity and belonging; formation of families, their reproduction and governance; and the relations of families to the wider world. Almost as soon as the team had decided on the three above themes, two of the research sites were badly affected by floods causing the research to be suspended. In returning to the sites, it was recognised that there was a need to look at the effects of crises such as floods on income and expenditure, community institutions and family relations. The flooding provided an opportunity to prolong and extend earlier short studies on migration and health and gave an important chance to explore crises of different kinds. The rationale for the health study mirrored that of the flood study (i.e. to examine the effects of health crises on household’s income and expenditure, community institutions and family relations). PERU The Peruvian team used the ‘thematic’ approach to explore the following themes:
Migrationwas of particular interest both because of its prevalence and because we expected migrants to have reflected on how movement affects life satisfaction. Variation in the nature and extent of collective action by different social groups was of interest because this can also have significant wellbeing effects, while the scope for collective action reflects group categorisation, goals and values. Research into consumption patterns was added in order to explore in more detail a key transmission mechanism from increased income to improved wellbeing. THAILAND The Thai team used the ‘thematic’ approach to explore the following themes:
This involved the study of the history and activities of particular groups that were active in the communities. The collective action research aims to explore the role of collective action in producing wellbeing at community, household, and individual levels by exploring who participates and who doesn’t, processes of inclusion and exclusion, motivations for participation, and processes of group creation and evolution. The livelihoods component investigates how changing livelihoods affect wellbeing in contemporary Thailand. It includes exploration of the ways in which changes in the economy and society are impacting on occupations, and the roles of migration and education. In particular, it explores the social and cultural construction of wellbeing through studying processes around changing occupations and changes in location (migration/mobility). It also investigates how changing resources and values affect livelihood strategies, and subjective and objective wellbeing. ETHIOPIA The Ethiopian team used the ‘core case’ approach to explore processes (this was called DEEP: the in-Depth Exploration of Ethiopian Poverty). The team used a mixed method and multi level approach (community, household and person) closely informed by the insecurity regime framework (Bevan, 2004).At the community level, protocols on community institutions and elites/destitution mapped the key institutions, organisations and individuals within the communities and their influence over processes of inclusion and exclusion. The young lives and old lives protocols explored the social and cultural construction of childhood/old age. Disputes and resolutions explore the concept of power through its actualisation in conflicts involving individuals, households and groups in different social contexts under study. Elements of household protocols were used to build a picture of the wider constraints, external intervention and shocks that affect communities (e.g. household wealth dynamics protocols). The migration protocol was developed to explore linkages between the rural sites with other rural and urban areas. The collective action protocol investigates the main purposes of collective action in the communities and identifies the different roles of internal and external actors. At the household level, monthly household diaries were carried out with members from 12 households selected from RANQ who were interviewed monthly about their activities in the previous months. This was used to produce a ‘combined household diary’ and ‘individual diaries’. Some of the protocols mentioned above were also carried out at the household level such as the migration and linkages module, the poverty and wealth dynamics module with an additional component focusing specifically on household shocks (its impact on resources, asserts and wealth and strategies to cope with and overcome shocks). At the person level, the team carried explored adult life histories looking at life courses, relationships and interactions across different life domains for 14 men and 14 women in each site. Further young lives and old lives protocols were carried out to understand the experience of childhood and old age. Additional research on individual experiences of migration for men and women in relation to marriage, inter-community social networks, trade, social services and other administrative and political purposes were also carried out. A protocol on intergenerational poverty dynamics expanded upon poverty dynamics protocols at household level by adopting a life-course approach to poverty and wealth transmission between generations. For more detail on the methods used and specific questions asked, see http://www.wed-ethiopia.org/deep_research_timetable.htm 6. How it was implemented The collection of data on ‘themes’ and ‘core cases’ was implemented by each of the four countries in the following ways. BANGLADESH Income and expenditure: The research drew upon three rounds of income and expenditure survey administered to 300 households in both rural and urban sites. This was complemented with one month’s process orientated research looking at structures and relationships of debt. The structures questionnaire was piloted, as was the survey. The decision to carry out research into debt structures and processes came about after an initial and quick analysis of round one of the income and expenditure, in which loans and debt were the most striking findings.Politics and community institutions: The focus of this research developed from the community profiles. Methodology involved mapping different institutional forms, interviews with members and non-members of the institutional forms, case study analysis and a final round of structured interviews focusing on a particular institutional type. In the end, it was decided to focus on political parties and mastaans since a) they emerged as particularly influential and b) our initial hypothesis is that they are linked. Marriage and family relations: Two rounds of data collection including life histories, in-depth interviews and focus group discussions in both rural and urban sites. There was a piloting phase before each of the rounds. Crises: Floods and Health: Both of these initiatives relied mostly on households diaries. 40 households kept diaries for the crises study. These were piloted and the questions were developed from our initial interactions and observations with flood affected households. We also asked 40 households in two of our sites to keep a year long diary in order to explore the dynamics of health seeing behaviour. This again was piloted and built on the initial study into health and health seeking behaviour. We supplemented the health data with specific questions added to the Income and Expenditure survey. This gives us data on health from across all our sites. PERU Social identity: Primary data collection on social identity was conducted by augmenting the WeDQoL questionnaire with an additional scale on social identity. This scale was an adapted version of one developed by A. Espinoza for his PUCP doctoral research into social identity of auto-rickshaw operators in Juliaca, Peru. Analysis of this data will primarily be statistical and integrated with that of other WeDQoL scales, as well as a personality scale adapted from the doctoral research of another PUCP psychologist, A. Calderón. Migration: The migration research was also organised in two phases. The first entailed semi-structured interviews with a target quota sample of two men and two women from RANQ households in each site in each of the following categories: migrant, non-migrants where family member had migrated out, non-migrant with no migrated family member, and return migrant. Persons fitting these criteria could not be found in each site, and as a result the final sample size was 71 out of a possible 112. Initial findings are available in an internal WeD document (Lockley, 2005). The second phase comprised a short questionnaire appended to the second round of the income and expenditure survey. Analysis and additional research on migration is being conducted by R. Lockley as part of her doctoral research. Collective action: The first stage of collective action process research involved the compilation of an inventory of all collective activities in each research site. The Peru team used ethnographic methods to construct a list of major activities affecting the wellbeing of members of households covered by RANQ that involve collaboration between them and/or outside agencies. Idealised accounts of what each was intended to achieve was contrasted with accounts of what was achieved in practice. The resulting inventory of forms of organisation informed the selection of specific case studies for the second stage. We chose to study one externally sponsored form of collective action (Glass of milk or Vaso de Leche) in order to compare how implementation of a supposedly uniform national program varied between sites. In addition we collected ethnographic data for one case study example of an informal collective action event, or faena, in each site. Vaso de Leche constitutes one example of a range of government social programs, in this case for provision of milk powder and other foodstuffs to expectant women and infants, through a network of local women’s committees supported by the local municipality. It can be seen as a progressive transfer to poorer people and women whose implementation is effectively decentralised. But it can also be regarded as a form of patronage that is prone to various forms of corruption and abuse. Our field team used participant observation and key informant interviews to complete a case study of one VdeL committee in each research site using a standard checklist of open-ended questions. The committee was purposefully selected to build on informal relationships they had build up and to overlap as far as possible with households covered by the RANQ and WeDQoL. This data was supplemented by a short opinion survey appended to the third round of the income and expenditure survey. A faena can be defined as pooling of voluntary labour for an agreed period in order to achieve a common purpose. Itis an institution with a long history as a form of community self-help in Andean society. A classic example is cleaning of irrigation canals. Such activities can be celebrated as a manifestation of collective culture and enlightened self interest. They can also be criticised as a form of labour coercion through which powerful individuals or agencies extract free labour from less powerful individuals. Our field team again used participant observation and key informant interviews to complete a case study of one faena in each site, using a standard set of open-ended questions. Each case study was purposefully selected to ensure the field worker could witness the event first hand, building on informal relationships they had built up previously and overlapping as far as possible with households covered by the RANQ and WeDQoL. Consumption: Consumption and subjective wellbeing has been researched by M. Guillen-Royo as part of her doctoral research. The research has been organised in two phases. The first consisted of 27 in-depth interviews and 4 focus groups carried out in a shanty town of Lima belonging to the WeD-Peru corridor. This phase explored priorities of consumption, motivations and institutions shaping choices in the slum. The second entailed the inclusion of questions about motives for consumption in the first round of the Income and Expenditure questionnaire carried out during July 2005 tackling 253 households in the seven communities of the corridor. THAILAND Health: In-depth semi-structured interviews were carried out with 24 individuals in each site, sampled according to gender, age, wealth status, and religion (in the South). A further 8 interviews were carried out with people with severe illness or disability, and another 2-5 with workers at the local health centre, Community Health Volunteers, and traditional healers. The interviews were preceded by grounding and piloting in the North East and South, and re-piloting in the NE after the interview schedule was revised following fieldwork in the South.Livelihoods: In-depth interviews were carried out at i) community and ii) household level, in both cases covering livelihoods, occupations, migration, and education. All households sampled had completed monthly Income and Expenditure diaries and interviews took place with household heads and individual members (migrant and non-migrant). The interviews were preceded by grounding and piloting in the NE and South, and exploratory interviews at key times for returning migrants like New Year and Harvest. The piloting led to the agreement of a final interview schedule at a joint workshop in Khon Kaen. Collective Action: Case studies were constructed in each village using 5-8 key informant interviews from groups selected using the community profiles and other secondary literature. These included a mixture of occupational groups (e.g. motorcycle taxi-rank and cattle raising groups), saving groups and religious groups. The key informant interviews looked at the group’s history, current status, activities, and the interactions between its members. They were supplemented by semi-structured interviews with members and non-members of these groups within the community (both those who had joined and left, and who had never joined). Questions covered group membership or non-membership, level of participation, activities of the group, importance to the community, personal benefits gained, problems encountered, relationships between members etc. The interviews were preceded by grounding and piloting in the NE and South, and exploratory interviews at key times for collective activities like New Year. The piloting led to the agreement of a final interview schedule at a joint workshop in Khon Kaen.
ETHIOPIA WeD Ethiopia used a range of different qualitative methodologies to investigate the ‘process protocols. These included in-depth semi structured interviews, focus groups, household diaries and many more dependent on the specific protocol. The research was carried out by twelve fieldworkers – 1 male and 1 female researcher in each of the 6 research sites; most were MA social anthropology graduates from Addis Ababa University. For most of the protocols, female researchers interviewed women and male researchers interviewed men in order to achieve a gender-balanced data set. The research was undertaken over a period from mid-July 2004 to the end of October 2005. Each month the researchers spent roughly 3 weeks in the field and a week in Addis Ababa reporting back, writing up and training for the next phase of work.Each researcher was given a fieldwork module which described the purpose of the research (sometimes including a conceptual framework); the content of the protocol; the number and type of respondents; a detailed list of topics/questions to be pursued; advice on ways to proceed, estimated time to be spent on the module; and a list of expected outputs. Often, the protocol was piloted before it was carried out in full. For more detail on the specific methodologies, see the guidance notes for each protocol at http://www.wed-ethiopia.org/deep_research_timetable.htm
The data produced from the process research in all four countries is mainly in the form of qualitative narrative-blocks of text. However, some of the survey elements of the research programme also offer insights into processes. The narrative data can be analysed both in itself and in relation to data collect by other elements of the methodology. Types of process questions the analysis can begin to answer include:
Details for how each country approached the analysis of their process research are elaborated below. BANGLADESH The analysis of each research component will depend on the research group leader. In working with the data, the intention is to let the data speak and develop themes, hypothesis, arguments from this. In some cases, technical applications are used in analysis (SPPS, NVIVO) while in other contexts we are working through various coding techniques. Again this will depend on the specific research topic.Initial analysis helps develop preliminary reports. We already have some for floods, health, mastaans, female employment, religious institutions, migration and quality of life. The next stage is to analyse further the data and incorporate it into key chapters. During this process some overarching arguments (in progress) will be identified and these will be fed back into subsequent drafts. A general principle of data analysis is that it should look at what, if any, relevant information can be found in the formal survey types methods adopted (RANQ, Income and Expenditure and Quality of Life). As far as is possible we want to integrate our methodological framework in our analysis. Finally, all researchers have been asked to keep two main issues in their minds at the moment of analysis:
PERU The qualitative data on Vaso de Leche was analysed using the questionnaire checklist to compare responses from each of the seven case study committees, supplemented with the results of the opinion survey (see Copestake, 2006). There is scope for further analysis of the opinion survey through cross-tabulation of responses with data collected using other instruments. The data on faenas was also analysed qualitatively, with particular emphasis on how their nature and extent varies along the urban-rural corridor that informed site selection, and its relationship to data obtained through Phase 1 QoL work in Peru on predominant goals and values in each site. This work is being conducted by J.L.Alvarez as part of his MRes(ID) at Bath. Initial analysis of the Phase 1 migration data has already been completed (Lockley, 2005; Lockley 2006). Social identity data will be analysed using quantitative methods along with other data collected on subjective wellbeing. Consumption data is being analysed by M Guillen-Royo as part of her doctoral research at Bath using a range of quantitative and qualitative methods.
THAILAND Preliminary analysis is taking place in Thai and English, and emerging themes will be discussed at a joint workshop in April 2006 to facilitate the drafting of workshop papers. Researchers are sharing their initial impressions by emailing around SPSS files, thematic coding reports from Nvivo, and short case studies.The data is currently stored in hard copy and MS Word files, although some analysis for Health is stored in Excel and Nvivo. Ultimately electronic copies of the original data sheets and preliminary analyses will be stored in the WeD-Thailand Qualitative database (MS Access), and will have English summaries to increase their accessibility to international researchers.
ETHIOPIA The analysis of the process data begins with an ‘immersion’ stage of all the data per site and per module. This allows a process of ‘learning by doing’ which will be supplemented by various workshops. Together, these various stages will allow a process of iteration which will allow specific themes to emerge. Specific analysis questions per module are detailed in the guidance notes.The protocol data gathered from the ‘core cases’ or ‘process modules’ have been inputted into a Word Database with a hyperlinked front page. This requires some anonymising, formatting, editing and ‘Englishing’ before it can be put into the public domain. The process research enables us to explain why certain outcomes are reached (or not) over others. Because different processes have different outcomes, direct links can made with several of the other WeD research instruments. For example, RANQ and Income and Expenditure survey gives information on objective wellbeing outcomes. These can then be compared with the QoL research which provides information on actors’ own perceptions. The process findings will also show how certain outcomes can influence actor’s goals and their resource profiles. Strong links can be made with the structures work since the whole process of generation of wellbeing and illbeing must be located in context. Social structures (encompassing cultural, economic and political structures) have a critical bearing on possible processes open to any actor. But they also influence perceptions of needs, resource distribution and the outcome of different processes. Altamirano, T., Copestake, J., Figueroa, A. & Wright, K. (2003) Poverty Studies in Peru: Towards a More Inclusive Study of Exclusion, WeD Working Paper 5 Bevan, P. (2004) “The dynamics of Africa’s in/security regime”, in Gough, I. & Wood, G. (eds) Insecurity and welfare regimes in Asia, Africa and Latin America: Social Policy in Development Contexts, Cambridge University Press: Cambridge, 202-254 Bevan, P. (2004) “Hunger, Poverty and Famine In Ethiopia: Mothers and Babies Under Stress”, WeD Ethiopia Working Paper 4 Bevan, P., Pankhurst, A. & Holland, J. (2006) “Power and poverty in Ethiopia: four rural case studies”, Paper prepared for the World Bank Poverty Reduction Group Copestake, J. (2006) “Multiple dimensions of social assistance: the case of Peru’s ‘Glass of Milk’ Programme”, WeD Working Paper 21, Davis, P. (2004) “Rethinking the welfare regime approach in the context of Bangladesh”, in Gough, I. & Wood, G. (eds) Insecurity and welfare regimes in Asia, Africa and Latin America: Social Policy in Development Contexts, Cambridge University Press: Cambridge, 255-286 Getachew, D. (2004) “Peasant Reflections on the Agricultural Development Led Industrialization (ADLI) Programme”, WeD Ethiopia Working Paper 3 Gough, I. & McGregor, A. & Camfield, L. (2007) “Wellbeing in Developing Countries: Conceptual Foundations of the WeD Programme, WeD Working Paper 19 Gough, I and McGregor, A. (2007) Wellbeing in developing countries: from theory to research, Cambridge University Press Gough, I. and Wood, G. (2004) Insecurity and Welfare Regimes in Asia, Africa and Latin America (Cambridge University Press: Cambridge) Guillen- Royo, M. (2006) ‘Are poor consumers getting what they need? An exploration of expenditures on health and autonomy in a Peruvian slum’ paper presented at the Global Studies Post-Graduate and Young Researchers Conference, Manchester, June 2006 Lockley (2005) WeD-Peru, migration and wellbeing study: initial findings. WeD internal working paper. McGregor, J. (2007) “Researching Wellbeing: from concepts to methodology”, WeD working paper 20 McGregor, A. & Kebede, B. (2003)“Resource profiles and the social and cultural construction of well-being, Paper to the inaugural workshop of the ESRC WeD research group (Jan 13th-17th, 2003) Newton, J. (2007) “Structures, Regimes and Wellbeing”, WeD Working Paper 30 Pankhurst, A. & Bevan, P. “Hunger, Poverty and ‘Famine’ In Ethiopia: Some Evidence from Twenty Rural Sites in Amhara, Tigray, Oromiya and SNNP Regions”, WeD Ethiopia Working Paper 1 Pankhurst, A. (2004) “Conceptions of and Responses to HIV/AIDS: Views from Twenty Ethiopian Rural Villages”, WeD Ethiopia Working Paper 2 Tadele, F., Pankhurst, A., Bevan, P. & Lavers, T. (2006) “Migration and Rural Linkages in Ethiopia”, Prepared for Irish Aid-Ethiopia Rural Case Studies” prepared for the World Bank Poverty Reduction Group
|
