Management and socio-economics of tsetse control

Quantifying the Costs of Living with Trypanosomiasis using a Questionnaire Survey

In some circumstances, a structured survey using a questionnaire may be the most appropriate way to quantify the costs of living with trypanosomiasis, and therefore the potential benefits of tsetse control. Some of these circumstances are:

• Where the difference between the costs of trypanosomiasis and the proposed costs of tsetse control as revealed by informal surveys is small, and within the margins of uncertainty of the informal procedure;

• Where there is concern that informal methods may not be able to cover a representative sample (e.g. may be biased in favour of wealthier or better-educated cattle owners), and where it is felt that a structured sample survey could in practice avoid such bias;

• Where it is felt that a structured sample survey could usefully explore differences between different groups of cattle-owners (or between cattle-owners and others) and;

• Where there is particular pressure from donors or governments for statistically rigorous results.

There are many sources of information on how to design a questionnaire and conduct a sample survey (such as Casley and Lury 1987). These refer to general questions of good questionnaire design, translation, pre-testing and analysis. The following are some areas where caution is particularly to be observed:

• If using household lists obtained from secondary sources, including government and donor offices, to construct a sampling frame, ensure that the lists do not systematically under-represent certain sorts of households (the poor, the geographically distant, seasonal migrants) or contain “ghost” households included in the lists to obtain benefits,

• Keep questionnaires short and focussed. Resist the temptation to ask about other topics on the grounds that the data may be useful later or for other projects. Participatory or informal surveys can help focus a questionnaire survey on information that is both significant and easily quantifiable.

• Be realistic about the capacity of informants, in particular:

To quantify information accurately.

To recall information accuratelyover long periods.

To distinguish between different animal health problems and different drugs.

• Be realistic about the capacity of interviewers to understand and accurately convey complex questions, to cope with conditional or branching questionnaires, and to calculate derived answers from informant responses,

• Be realistic about the time and human resources necessary to clean, enter, analyse and interpret data.

Click here to download an adapted version of a short, focussed questionnaire offered for the purposes of illustration, covering:

• Overall cattle numbers and herd composition

• Mortality due to trypanosomiasis and other causes

• Costs of treatment

• Ploughing days lost