«Juanwen Yuan Thesis committee Thesis supervisor Prof. dr. A. Niehof Professor of Sociology of Consumers and Households Wageningen University Thesis ...»
In March 2008, I continued my life history recording and case study interviewing. I also began to input the survey data in the computer and to process the qualitative data. In April 2008, it was sowing time for the villagers and I did not go to the field, but continued the data input process. In May 2008, I carried on with my life history recording and case study interviewing. My promotor visited the research site, discussed my progress, and provided suggestions for the remainder of the fieldwork. We also planned the detailed schedule for the next four months.
During June-September 2008, I tried to find some secondary data I missed and held key informant interviews with former village leaders and elderly people.
I conducted 14 focus group discussions: two mixed group discussion; ten female group discussions; and two male group discussions (see Section 5.3.5 for the topics discussed and the composition of the groups). I also finished incomplete interviews and case studies and checked for missing data. These were the toughest months for me to finish verifying my questionnaire because I found some data were missing, something I did not check very well in the survey period. Participant observation was integrated in the whole process. I reflected on my data, did the coding of data, and carried out some qualitative data analysis.
5.3 The data collection I began my formal data collection after I went back to China, in July 2007. As I mentioned before, participant observation was an important method for me to collect data in the field, since I had already been working in the research area for ten years and the local people had truly accepted me. They were not easily disturbed by my presence, even when I brought along strange company. They would greet me and continue to do what they were doing. I held many informal interviews with them, in which they would sometimes give me very confidential information. I took field notes almost everyday. I also could compare the situation with that of ten years ago, when I worked in the area on a project. I lived with the local people and could observe their daily lives, the division of labour in household work and agricultural production, animal husbandry, marketing, weddings, and other social events. I accompanied the women to look after cattle grazing in the hills. I went to the public spot under the big tree for social village activities, where villagers get together to chat and outside traders come to do business.
Even though I was very familiar with them, I still tried my best to make them comfortable and not feel disturbed when I recorded my conversations with them. I found it worked well to use the cell phone to record interviews and discussions. The merit of this method is that the interviewees do not notice that you are recording and will not be distracted by it. After the interview and discussion finished, I told them that I recorded the interview by cell phone and solicited their feedback. None of them minded my recording the conversations, but some said that their lives were not interesting or important enough for me to record their stories so carefully.
69Study design, data collection and analysis
5.3.1 The household survey The household survey was conducted in seven villages. The questionnaire (see Annex 1) focused on obtaining information about the current situation of the households. I mainly investigated household resources, livelihood activities and gender roles. I included 160 households in the household survey16. In addition, 24 migrated households (see above) were interviewed as well about their motivation for migration, their plans for the future, the work they were doing, and the management of their land in the village. The interviewees mainly were household heads.
Sampling As mentioned in Chapter 4, the municipality of Kaizuo in Guizhou province was chosen as the research area. Agro-ecological cluster sampling was conducted first, after which the sample was stratified according to cohort in each agro-ecological cluster. From an agro-ecological perspective, there are two kinds of villages in the municipality of Kaizuo: upland villages and lowland villages with paddy fields.
The agro-ecological differences are important because the division of labour, migration, land use, and livelihood strategies are different in these two ecological systems. The upland households mainly grow maize, while those in the lowland area grow rice. The lowland households have relatively easy access to markets, while the upland households have not. Water is relatively abundant in the lowland area and very scarce in upland areas. The level of education is higher in the lowland area, because the villagers can access schools easily. The lowland households use coal as the main source of energy, while the upland households use fuel wood. The households in the lowland area have better access to government development projects, while those in the upland areas have more opportunities to get support from poverty-alleviation projects.
I selected 160 households: 80 in each agro-ecological cluster. In addition, I also surveyed 24 migrated households. Seven villages were selected for the survey, as explained in Chapter 4: Dabuyang, Guntang, and Xiaobuyang in the lowland area, and Dongkou, Xiaozhai, Xingzhaiyuan, and Dabang in the upland area. The main research villages were Dabuyang and Dongkou. The other five villages were added to complete the sample. The numbers of households sampled were 49, 41, 41, and 29, respectively, for the 1970s, 1980s, 1990s, and 2000s cohorts. Based on the design, there should be 40 households in each cohort. I tried to balance the number of households in each cohort, so as to make the data analysis more powerful.
However, it was difficult to find 40 households involved in agriculture for the 2000s cohort, because there is a lot of migration in this cohort. Hence, only 29 households represent this cohort in the survey. The 1970s cohort is represented by 49 households. The details of the composition of the sample are provided in Table 5.2.
160 households include the 10 pre-survey households, but I had to get some answers 16 which were missing in the pre-questionnaire.
During the survey, I found some couples who had migrated in the past year and had come back for the Spring Festival. This raised the issue of the characteristics of such migrated households17. To compare them with the other households, I randomly selected 24 migrated households: 5 households from the 1980s cohort, 8 households from the 1990s cohort, and 11 households from the 2000s cohort. No migrated households were included in the 1970s cohort. The questionnaire can be found in Annex 2.
Training and working with enumerators I spent three days to train one enumerator in advance, because he was available.
We did the pre-survey together. One week later, the other four enumerators were trained and joined the survey. The enumerators were majoring in rural regional development and all had rural experience. After I trained them, I asked them to do the pre-survey and clarified some questions. Every evening, we shared the survey results and reflected on these, while we sat in front of the fire stove to get warm.
We found the discussions to be very helpful for the survey work the next day. We reached agreements when there were different understandings about the survey questions.
5.3.2 Secondary data collection I already had some secondary data. Before I collected the rest of the secondary data, I went through those I already had and collected additional materials on local agricultural activities, government policies, agricultural extension mechanisms, and recent data on the municipality’s socio-economic status. I also got the data from the Second Chinese Agriculture Census, finished in Kaizuo in 2007. I studied these materials before I conducted the survey to obtain a clearer picture of the seven sampling villages, and I also used the materials to check the reliability of my survey data.
A migrated household means that both husband and wife have migrated with or 17 without their children.
5.3.3 Key informant interviews I held 31 key informant interviews. I interviewed three government officials, two (former) extension workers, 15 (former) village leaders, three (former) female village leaders, one unofficial village leader (zhailao), and seven elderly people. I selected the municipal officials who had been working in the municipality for many years. All these interviewees told me about the current situation. Some told me about the situation in the past and during the collective era. I also conducted interviews on the local culture with these interviewees. I conducted key informant interviews individually, but sometimes two or three people attended the other interviews. The detailed interview outline can be found in Annex 3.
5.3.4 Self-ranking I did self-ranking twice, once in Dongkou and once in Dabuyang. Villagers got together during my interviews and if more than five participants were present, I seized the opportunity to have them perform a self-ranking with regard to their socio-economic status. The results of the self-ranking helped me to understand the heterogeneity in household status. The villagers ranked the households based on income, housing, livestock holding, trucks, and landholding. They did the ranking with the aid of an integrated evaluation of the above indicators. I had people ranking themselves in the two main research villages (Dongkou and Dabuyang).
Three levels of household wealth were distinguished in the self-ranking discussions: poor, middle, and rich households. In the survey, I also asked each household to evaluate its economic position in the village according to the five categories given in the questionnaire (see Annex 1).
5.3.5 Focus group discussions (FGD) I held 14 focus group discussions (FGDs). Focus group discussions were organized for each cohort in the two villages. Each group should preferably include no more than 12 persons ((Krueger and Casey, 2000). In this research, each FGD included five to seven persons. The sessions usually took half a day. Four FGDs were organized for male groups in the 1970s cohort (cohort 1) and the 1980s cohort (cohort 2) each, and four FGDs were set up for female groups in the same cohorts.
Some participants were randomly selected within the same cohort, some after consultation with village leaders because of the rich information they might provide. We discussed the changes they experienced in the different stages of their life course. The 1970s and 1980s cohorts discussed livelihood, land use, food security, and gender issues in both the collective era and the HRS era. Four FGDs were held with the 1990s (cohort 3) and 2000s (cohort 4) cohorts each. With these participants, their livelihood, land use and gender issues in the HRS era were discussed. Two FGDs with mixed people of different ages and gender were conducted in Dabuyang village and Dongkou village to compare opinions of men and women of different ages with regard to the process of social change during the past 30 years. The issue of female-headed households was discussed in this group as
well. Mixed groups are not easy to manage ((Krueger and Casey, 2000), but I tried my best to moderate and avoid conflicts in the discussions. Table 5.3 lists the topics of the FGDs.
5.3.6 Life history I documented the life histories of eight women of the 1970s, 1980s, and 1990s cohorts (see Table 5.4). I selected them randomly in the villages of Dabuyang and Dongkou. I directly went to their houses and checked whether the woman had time or not. If she had time at that moment and would like to talk to me, I stayed with her and conducted the interview. I found most of the women to be generally open and talkative. The fact that I had already communicated a lot with the women in the villages at the time I was involved in a rural development project there, probably helped to gain their trust. They liked to tell me their stories and share their experiences with me.
In order to better understand the different households and have in-depth discussions about the life histories, I chose four households of different cohorts to stay with for a shorter or longer period of time: one household of the 1970s cohort and three households of the 1980s cohort. I also shared meals with many households and tried to talk freely with the women during meals.
Conducting life history interviews is time-consuming. I had to go back to the same interviewee several times to finish the whole story. Some interviewees talked a lot and seemed interested in telling me everything. I found it difficult to cut their talks short and tried my best to draw them to my topics. I also did some preliminary analysis after I finished an interview, to prepare the follow-up interview with the same person. Some husbands also joined in the interviewing process and gave additional information, but this did not happen often.
5.3.7 Case study Seven case studies of households (including migrated households) were conducted in three villages, apart from the eight life histories: three for the 2000s cohort, two for the 1990s cohort, one for the 1980s cohort, and one for the 1970s cohort. Of these seven households, four households are from the village of Dabuyang, two from Dongkou village, and one is from the village of Guntang. They were selected for
74 Chapter 5