An inventory of fish farmers to update records was conducted by the Department of Fisheries prior to the survey. When the survey team arrived, the Department of Fisheries records were still under preparation. Kawambwa was the only district with complete records. It was agreed to start the survey at once in this area, while work on updating records continued in the remaining districts. The Integrated Rural Development Programme (IRDP) provided a vehicle for this purpose.
From the records the number of fish farmers in the province was estimated to be around 270. It was agreed that 25% (67 fish farmers) had to be interviewed to obtain statistically reliable information. Furthermore, it was planned to interview 25 potential fish farmers, giving a total of 92 interviews.
The sampling was done on an individual level and had to be done separately for each district as the records to start with were incomplete. Random sampling was performed by picking every fourth farmer from the records. Individual sampling results in a lot of travelling. But so would village level random cluster sampling, as in most cases there was only one fish farmer in each village.
It turned out during the survey that the total number of fish farmers on the records was 311. To some extent these records included potential (intending), practising and former fish farmers. Sampling every fourth from this number resulted in 78 random samples. The discovery of potential (intending) fish farmers in the records, and their inclusion in the sample was not anticipated. There was less need to approach potential fish farmers in the course of the survey. The sample was still large enough for the target of practising/former fish farmers, as it became apparent that around 30 of the farmers, listed by Department of Fisheries as fish farmers, did not exist. One quarter (25%) of the existing 280 farmers gave a target number of 71 former and practising farmers.
Samfya District was an exception to what is stated above. The records for this district contained so many potential fish farmers that additional sampling had to be done to achieve a sufficient number of practising/former fish farmers.
It was decided, to the extent possible, to carry out 100 interviews (50 practising, 25 former, 25 potential) to ensure as good a statistical validity as possible.
At the planning stage, the consultant agreed with ALCOM staff that the survey staff would strive to meet certain conditions for the interviews:
Each farmer should be interviewed in the presence of both an aquaculturist and a person trained in social sciences. The fish ponds should be inspected by the aquaculturist;
Fish farmers should be interviewed alone without neighbours, friends and other fish farmers present. They could then feel free in expressing attitudes and providing answers about their activities;
Before starting an interview, the enumerator should carefully explain the purpose of the survey, and the type of questions to be asked. The farmer should be given the opportunity to ask questions before and after the interview.
It proved impossible to meet all these conditions during all interviews. In most cases, it was not possible for the social scientist and the aquaculturist to be both present during interviews, though they worked closely during the entire survey.
The third and fourth conditions posed no problem. However, Department of Fisheries staff were present as interpreters during a number of interviews. In others, the Department of Fisheries staff member accompanied the fish culturist to the pond site and the social scientist was then the only survey team member present during the interview.
The survey was executed in two visits during June-August 1988.
During the first visit (20–29 June 1988) the survey was conducted in Kawambwa District. The first interview was carried out on the 22nd of June and the second on the 25th of the same month. 32 interviews were carried out (20 practising, 3 former and 9 potential).
The survey team included Ms. Lilian Chinkumbi, social science graduate from the University of Zambia, and ALCOM staff member Mr. Karl Otto Wahlstrom, Aquaculturist (APO). From the province Mr. Cyprian Tembo, Fish Culturist in Mansa, participated.
It was possible to form only one survey team and there was only one vehicle available. It was essential for the Department of Fisheries Fish Culturist to take part in the survey team as he was the person who could locate existing fish farmers.
The interviews were carried out by Ms L. Chinkumbi while Mr. Karl Otto Wahlstrom and Mr Tembo visited the ponds. When any additional sampled or potential fish farmer was available for interview in the vicinity, he was interviewed by Mr Wahlstrom. Where the respondent's knowledge of English was insufficient, Mr Tembo acted as interpreter.
Though it was preferable to interview the fish farmers alone, fisheries staff were present as interpreters. Circumstances and time pressure made this arrangement necessary. It is the survey team's opinion that when an interpreter was necessary, the farmers did not consciously bias their answers.
During the second trip (25 July – 11 August 1989) the survey was conducted in Mansa, Mwense and Samfya Districts. The first interview was carried out on 26 July and the last on 10 August. During this period, 62 interviews were carried out (31 practising, 17 former and 14 potential fish farmers).
The survey team consisted of Mr Wahlstrom and Ms Nelly Mazingaliwa, social science graduate from the University of Zambia. The team was assisted by the following Department of Fisheries personnel from the province: Mr Cyprian Tembo (Fish Culturist, Mansa), Mr Mubumba (Fish Scout, Mansa), Mr Mulenga (Laboratory Assistant, Samfya) and Mr Robinson (Fish Scout, Samfya).
On the whole there were no major difficulties in selecting farmers at random, except in Samfya District where some nearby farmers had to be interviewed as the initially sampled farmer(s) could not be found. In Samfya District there was also a need to sample a few additional practising and former fish farmers as the records, it turned out, included many potential fish farmers.
In general the interviews were carried out by Ms N Mazingaliwa, while Mr Wahlstrom and staff from the Department of Fisheries visited the ponds. Where additional sampled or potential fish farmers were available in the vicinity, interviews were carried out by Mr Wahlstrom as during the first visit.
Most sampled fish farmers were found and interviewed by “chance”. The majority were found around their houses, some were located in the fields or at the ponds, while others were fetched by household members. In quite a number of cases members of the household told the respondent to come to the house. In the meantime the survey team proceeded to the next respondent, and returned to carry out the interview.
Survey coverage
The following table indicates the extent to which the survey plan could be followed. It shows the number and location of farmers the team was scheduled to interview (in accordance with the random sampling) and the farmers actually interviewed.
Table 2.2: Interviews with farmers -- as planned and as executed
| INTERVIEWS WITH FARMERS AS PLANNED | |||
|---|---|---|---|
| DISTRICTS | POTENTIAL FARMERS | PRACTISING AND EX-FARMERS | TOTAL |
| Kawambwa | 9 | 26 | 35 |
| Mansa | 8 | 22 | 30 |
| Mwense | 4 | 11 | 15 |
| Samfya | 5 | 11 | 16 |
| Total | 26 | 70 | 96 |
| FARMERS ACTUALLY INTERVIEWED | |||
|---|---|---|---|
| TOTAL | PRACTISING FARMERS | EX-FARMERS | POTENTIAL FARMERS |
| 32 | 19 | 4 | 9 |
| 33 | 22 | 7 | 4 |
| 15 | 5 | 5 | 4 |
| 14 | 5 | 4 | 5 |
| 94 | 51 | 20 | 23 |
Thus, the team achieved its target with respect to interviewing practising and former farmers (71 as compared to 70). Two more potential farmers should have been interviewed but it was difficult to locate any while finishing the survey in drought-struck Mansa District.
Questionnaires and their analysis
The survey questionnaires were designed by ALCOM. They had been tested during previous surveys of the Northern Province and subsequently modified. For more details see reports FI:GCP/INT/436/SWE.4 and FI:GCP/INT/436/SWE.5.
The questionnaires are designed to solicit information which describes the farmers' current situation in terms of: (i) the pond (ii) nature, quantity and use of inputs (iii) harvesting practices (iv) use of harvested fish, and (v) the farmers' problems as he reports them.
To establish the identity of the “fish farmer” the respondent is subsequently questioned about: (i) his motives and, the extent to which they are circumscribed or constrained by social or cultural factors; (ii) his social, and economic situation; and, (iii) allocation of resources amongst fish farming, other farming and non-farming activities.
These data, when analyzed against the “current situation” information, will indicate the degree to which individuals with different backgrounds are: (i) more or less constrained in their management of fish ponds (ii) likely to continue or abandon the activity; and, (iii) the degree to which pond management practices might be influential in determining success.
The data obtained during the survey were entered on a data base management programme operated on a personal computer. The same programme was used to analyze the data.
