A survey performed by the Department of Fisheries in Northern Province in early 1986, covered 731 fish-farmers with a total of 1 992 ponds, distributed in 276 villages and 28 chief areas. This information made possible the sampling procedure and proved to be of great help when locating the selected fish-farmers.
The initial objective of the survey was to interview a total of 100 practising and former as well as 30 potential farmers during a 3-week period. For the sampling it was decided to apply a random single stage cluster sampling model where the villages were defined the cluster units. The teams would thus spend less time travelling between farmers. All the villages in Northern Province were numbered; from 1 to 276 and 50 random numbers in this interval were generated. This gave a sufficient number of interviewees and also some backup in case the respondent was not available. The potential fish-farmers were identified either by the fact that they had recently contacted the Department of Fisheries staff for start-up advice, or had asked existing fish-farmers for suggestions.
Two four-wheel drive vehicles were made available for the survey, both kindly lent by NORAD. This made it possible to travel in 2 teams. Fourteen days before the actual start-up of the field work, 4 enumerators were selected; 2 aquaculturists to cover the professional technical knowledge and 2 sociologists for the economic/sociological questions. Three team members were able to communicate well in the local languages of the Northern Province. The complete survey team gathered in Lusaka 3 days before departure for the field to study the questionnaire and learn the instructions/objectives of the different questions. A preliminary survey schedule was arranged and preparations made for a 3-week stay in the field.
The survey teams could not locate the fish-farmers without assistance. In Kasama District the teams were accompanied by fish scouts from Misamfu Provincial Fish-Farm. In the other Districts these scouts very rarely had vehicles and therefore were not sufficiently familiar with fish-farmers. In Mporokoso and Mbala Districts the teams were assisted by the agriculture camps and their extension staff. The majority of the fish-farmers are also agriculture-farmers and regularly receive visits from the camps. These extensionists usually also had some information on their fish-farming activity, and their problems and needs.
For the first 4 days the enumerators travelled in one car. If the farmers were situated near each other, they divided into two teams each with one aquaculturist and one sociologist. The car was moving between the teams bringing them to new farmers. When the second car arrived, the enumerators were permanently split up into two teams. Initially, the aquaculturist travelled in the first car and the sociologists in the second, as was suggested before the survey. This arrangement, however, proved problematic. First, it was very difficult for the aquaculturists to inform the sociologists about which farmers were encountered and available for interview, and exactly where they were located. Second, some farmers felt that two separate visits and interviews took too much of their time. It was therefore decided that each team should have one aquaculturist and one sociologist. In the mornings, before departure for the field the teams agreed upon which farmers were to be visited by each team, taking geography and distance into consideration. This procedure worked well.
When the respondent was not at home, an attempt was made to make an appointment for a later meeting. If this was impossible, the teams in a few instances interviewed another member of the economic unit, if this person appeared to have reliable information about the fish-ponds and the fish-farmer's activity. If neither of these alternatives were possible the back-up list was checked to see if there were other selected farmers in the village. They were then contacted for an interview.
It proved to be impossible to reach the intended number of interviews. During 13 working days in the field, of which 4 days with one vehicle only, a total of 89 interviews were conducted on 46 practising, 22 former and 21 potential fish-farmers. This gives an average of nearly 7 interviews per day. For the practising and former farmers the interviews lasted for 1–1 1/2 hours including inspections of the ponds, provided they were situated fairly near to where the farmer was encountered. For the potential farmers the interviews lasted for 15–45 min.
The questionnaire was constructed to facilitate the treatment of the data on the computer. Most of the questions offer yes/no or multiple choice answers. A computer data base package was applied to enter the coded data and later make the analysis. The data entry system was developed by a data specialist at FAO Headquarters and completed by Programme staff.
The actual analysis was done by the authors in Lusaka on a computer provided with a 640 k memory and two disk drives. In addition to the authors, the other three survey team members had the opportunity to comment on an early draft of the report on the survey results.
