TOOLS AND DATA COLLECTION IMPLEMENTATION

A keen selection of the primary data collection tools was done in order to allow for triangulation of the information collected. Key informants’ interviews at both field and headquarter level, observations and secondary data tools were used to complement household interviews as follows.

3.2.1 Desk review of existing literature

An extensive literature review was collected that focused on the susceptibility to HIV and vulnerability to AIDS as well as the interface between Food security and HIV/AIDS. Internet

19The head of the homestead alive or deceased may have a family composed of one or more wives and their children. The elder sons are also married and living in the same home making an intertwined network of households. These greatly differ from one to another.

sites were also visited to obtain specific data on the research issue from available papers.

Reports from the government and organisational documents were used. Local news papers such as the Kenya Daily Nation, was also instrumental in providing information on the extent of the food security crisis situation that had created an alarm in the whole country hence inspiring the cabinet to develop an emergency action. All this data was instrumental straight from the research proposal development to the data analysis whereby the available literature was compared with the research findings in the context of this research.

3.2.2 Qualitative methods

Qualitative methods used are household and key informant interviews, FGD and observations. The household interviews took great care to respect confidentiality and all data was kept anonymous except for few informants who were quoted by their designation.

a) Household interviews

Household interviews were conducted through field visits. Field visits were conducted to Mkenge, Msabaha, Dabaso, Jimba in Malindi district and Shononeka in Kilifi district in the Coast province. These interviews used a semi-structured questionnaires (Refer to Annex 12) to explore the demographic characteristics; assets base; trend of food availability; ways in which households were coping with food insecurity; the constraints faced as well possible interventions. The household questionnaire was administered to the most knowledgeable member of the household who in most cases was the father or mother of the household.

Interview informed consent was sort from the respondents prior to administering the questionnaires in each household. A total of 22 interviews were conducted in the two categories. These interviews served to cross-check and validate interpretations from the secondary data and to make clear some common facts and issues. With these basic purposes in mind, and given some that respondents’ desired not to be quoted, individual interviews are not directly referenced in the report.

The use of closed questions in the semi-structured questionnaire made it easy to gather responses from the interviewees and assist in probing suggestions whenever they were stuck.

b) Focus Group Discussions (FGDs)

Focus Group Discussions were conducted in both Malindi and Kilifi districts where both male and female were involved (Refer to Annex 15, photograph 2).

The participants for the FGDs were randomly selected to combine men and female representatives from the two locations of Gede and Watamu in Malindi district. In Kilifi they all resided in the same location. A total of 8 FGDs (4 in Malindi and 4 in Kilifi district) were conducted, where the male and female jointly participated in the sessions apply the four tools. This tricky situation in both the two districts was as a result of having more female participants during the FGDs as compared to male participants (only 2 male participants in Malindi and 3 male participants in Kilifi). This is because some would promise to participate during the planning only not to turn up on the material day due to unavoidable household responsibilities. Even though this was a very big challenge in data extraction for gender specific information which may translate to biasness of information extracted during the analysis. Coincidentally, there was active participation of all sexes since the male were few so the female felt secured to air their views openly. They would at times clash at times complement each other when providing gender specific information. They participated throughout the entire sessions since the time was pre determined at the start of the sessions allowing for flexibility. This though was counteracted by making sure attention was given to

each gender during the discussions. For purposes of triangulation of results, the male respondents were specifically targeted as respondents during the household interviews.

Another limitation with this tool was that it combined ideas of both the rich and the poor and yet critical issues of the rich and poor may have different aspects, perceptions and these may be very important for planning and developing interventions that will address food insecurity.

These discussions were facilitated making use of four Participatory Rural Appraisal (PRA) tools not only to collect information (UNAIDS and KIT, 2004) but also to discuss solutions as highlighted below (For details refer to Annex 13).

i. Seasonal calendar: This tool was applied to show the changes in activities of the people during the different seasons in a given calendar year. This is used to gain insight of people’s time spending, movements away and back and compare with food availability so as to identify the specific periods of risk.

Below is a combination of tools that were used to complement each other in facilitating participants to identify the risks of HIV infection through the adopted coping strategies.

ii. Appraisal of risk behaviour: This tool was used to help participants identify different types of risky behaviour or conditions that may lead to risk behaviour predispose them to HIV infection. It was purposely selected to increase awareness that there are different risk behaviours and that many people are at risk to become infected with HIV. Solutions were also discussed to address the different risky behaviour and which people.

iii. Mapping of risk areas: This tool was used to identify places, environments and analyse situations of risk for contracting HIV in the community and map them. The purpose of this tool was to identify where sexual risk behaviour takes place, where it is negotiated or people feel at risk of contracting HIV.

iv. Occupational risk: This tool was used to compliment and to help participants to discover the risks for HIV infection of different people in different occupations.

c) Key informant interviews

The 5 key informants interviewed were conducted. The interviews were limited to district headquarters, and largely and largely the heads of the office (See Annex 14). They were strategically chosen because of their positions and their stake in food security responsiveness in the area of study. These provided information on: the coping strategies of the smallholder farmers in the study areas, influences for those decisions taken by the smallholder farmers, available safety nets within the community, current and possible interventions to help solve the problem of food insecurity currently facing the smallholder farmers (Refer to Annex 15).

d) Observations

Observation was mainly done for ‘people watching’ and ‘situation watching’ during field work.

‘People watching’ was where the interviewees and those around were observed for their state and condition. Whereas ‘situation watching’ involved observing the study area for the position and extent of food insecurity in relation to the study topic.

The observations were achieved in the homes during the questionnaire administration at household level; and cross-cutting the markets, trading centres and villages when moving

from one village to another (Refer to photographs in Annex 16). This was used to explore more on the extent food insecurity faced by the farm families and on the coping strategies of the smallholder farmers’ households. Triangulation of information made it possible to validate what the respondents said and what is observed. Some clarifications were provided by the village elders where possible through question and answer.

Own experiences and that of others was also the basis used to describe the context of the problem so as to derive the problem statement and develop the semi-structured questionnaire and checklist to be used in data collection. This is also the basis of selection of the sample area and key informants.