Methods of summarizing and presenting RA results
Some sources of bias from RA techniques
Conclusions
Bibliography
M. Ghirotti
The author is Animal Health and Production Officer of the Central Technical Unit, General Directorate of Cooperation for Development (DGCS), Italian Ministry of Foreign Affairs, Via S. Contarini 25, Rome, Italy. He wishes to thank Dr S. Sandford, International Livestock Centre for Africa (ILCA), for sharing his vast experience with the Ethiopian livestock systems that were the first test bench for the methodology described in this paper, Mr A. Carloni, rural sociologist FAO, Rome, and Dr R.B. Griffiths, former Director, Animal Production and Health Division,
FAO, for their suggestions and constructive criticisms to the script, and Ms S.Z. Babsa. An earlier version of this paper was presented at the tenth meeting of the Society for Veterinary Epidemiology and Preventive Medicine held at the University of Edinburgh, Scotland, 1-3 April 1992.
This paper describes rapid appraisal (RA) methodologies, which aim at obtaining a quick, systematic and cost-effective picture of livestock conditions and veterinary problems, particularly in agropastoral communities, and involving farmers in the process.
The original methodology was first developed in 1985 in the Ghibe Valley in the central highlands of Ethiopia. It was further tested and improved in various areas of the Shoa region in Ethiopia, as well as in the Kafue flats of Zambia, the Quiché area of Guatemala, the province of Sidamo in southern Ethiopia, Owamboland of Namibia, the Boké district of Guinea and the Nouhao Valley of Burkina Faso. It includes techniques often associated with standard surveys used to confirm qualitative information through the recording of selected parameters and indicators. The basic conceptual framework is shown in Figure 1.
Animal health and production interventions can be considered as part of the process to develop a given area and community. Availability of appropriate data is crucial for the planning, management-and evaluation of such activities so that resources-can be best invested and intended results achieved. Although epidemiology has provided powerful tools for this purpose (Abramson, 1984; Putt et al., 1987), too often the collection, consolidation, elaboration and feedback of data obtained through conventional surveys is time-consuming. Moreover, in many developing countries background information is scarce, communication networks poor and resources very limited.
There has been increasing concern among researchers and development workers of different disciplines about the need to develop and utilize field diagnostic techniques that, while trying to combine rapidity with sound scientific methodology, allow for the systematic identification of constraints, the testing of possible solutions to overcome them and the monitoring of activities. Decision-making has often been subject to the spatial. social and seasonal biases of "development tourism" (visiting project areas near main roads during quick trips and contacting local representatives of privileged segments of the society in healthier periods of the year), to the lack of a common framework for data collection and to the sectoral approach of most academic researchers (Carruthers and Chambers, 1981).
A major contribution has been made by the social sciences and by the "participant observation" approach, which involves the direct study of a community and, as much as possible, participation in its life and activities (Edgerton and Langness, 1977). Community involvement is widely considered to be a critical factor in the success of any animal disease control and livestock production programme, as it is essential to the implementation as well as to the planning and evaluation of any developmental activity (ILCA, 1977; WHO, 1990).
The health and productivity status of a livestock population should be analysed within the framework of its farming system to try to understand the role played by the different components. Modern and traditional veterinary care should be considered as part of the whole health system (Schwabe, 1978; Jahnke, 1982; WHO, 1990). Disease prevalence or herd structure and productivity are dynamic processes that result from the interaction of different factors, both internal and external to the system. The nature and respective importance of such determinants may not be fully appreciated by an external investigator alone, especially when the causes are particular to the local situation, where the key to interpreting raw data may be provided by the farmers' experiences.
Standard questionnaire surveys with fixed choice responses do not give beneficiaries the opportunity to discuss potential project objectives and activities. Over the last ten years, RA techniques have been developed to benefit from farmers' experiences and perspectives. RA has been defined as "a systematic. semi-structured activity carried out in the field by a multidisciplinary team and designed to acquire quickly new information on, and new hypotheses about ... development" (McCracken, Pretty and Conway, 1988).
Steps in rapid appraisal
The main steps in carrying out RA for livestock development follow.
Preliminary identification of problems and RA objectives. As suggested by McCracken, Pretty-and Conway (1988), there are three different types of RA: exploratory, monitoring and topical. RA may be used by veterinary and livestock officers or by researchers to analyse and describe the general features and constraints of livestock production systems present in the area where they are working (exploratory RA). This is recommended for the establishment or redefinition of development programmes and for the organization of local livestock services according to farmers' needs. RA may also be performed to evaluate the progress or the impact of development activities or veterinary services (monitoring RA). In both cases it is important to clarify the target group since livestock ownership and husbandry practices differ among ethnic and social groups. For example, in many societies low-income households seldom own large animals but instead raise small ruminants, pigs and poultry. Moreover, the function of livestock, herd composition, feeding practices, productivity and problems of livestock found in underprivileged social or ethnic groups are different from those apparent in other segments of the society. Examples of factors that may differ include accessibility to drugs or to supplementary feeding and herd composition (Ghirotti, 1988). If such general information is already available, RA may be used to select more specific objectives and areas of investigation according to the hierarchy of needs expressed by the community and to research objectives (topical RA).
Additionally, we suggest that researchers could carry out RA together with investigations on already identified topics to obtain a clear conceptual framework for a better interpretation of the results of more selective studies (framework RA). For example, the RA methodology was tested as a diagnostic tool to identify and describe local livestock production systems in the Ghibe Valley area at the same time as the influence of grazing dynamics on the reproductive behaviour of local zebu-cattle and on the role of male animals in herd fertility was being investigated (Ghirotti and Woudyalew, in press). In southern Zambia, RA was utilized to identify husbandry practices and cultural and environmental factors that may help to interpret the results of a sero-epidemiological investigation into major cattle diseases and to identify valid and acceptable control measures (Ghirotti et al., 1991.)
Constitution of the team. The team should benefit as much as possible from the contributions that different disciplines can provide in the management of animal resources. It should not consist of more than four persons, however. A small team is probably more agile, discreet and likely to be accepted by the community during field visits. The division of duties and the sharing of opinions is easier and faster among fewer members.
Preferably, the team should be composed of a veterinarian with sound knowledge of animal production, epidemiology and disease economics, an ecologist/agronomist with a good background in pasture management and animal population dynamics and a social scientist acquainted with pastoral and agropastoral societies. Each person should have an intersectoral approach to problem analysis and solving.
The composition of the team can be modified according to available human resources and to the nature of the specific topics to be investigated once the presence of members who are competent in the above-mentioned disciplines has been assured. Local technical officers should be included in the team. As the team is formed, the leader should organize a workshop to discuss with the other members the study objectives and methodologies to be adopted, including criteria for the selection of the communities to be surveyed.
Identification of target areas and communities. The choice of target areas and-communities mainly depends on the study objectives, the type of RA selected and the available resources.
The population involved in the analysis should be stratified according to the selected variables to be investigated, such as: main agro-ecological and production systems, for example, intensive as opposed to traditional livestock systems and different ecological systems; the presence or absence of a given factor, for example, crossbreeding schemes, vaccination, tick control and other veterinary services, and credit facilities; and different sociocultural groups. The target communities can be those involved in a given development programme that has been implemented in an a priori defined area. When analysing the general features of a given farming system or monitoring a specific programme with limited resources, such as time, funds and transport, the selection can be based on "representative" reachable villages. In some countries or regions permission to travel or contact with farmers may be restricted for security reasons or because the communication network is poor. In such cases, the team should be aware of the potential sources of bias and of the difficulties in making generalizations based on the results of an analysis thus carried out. If the above-mentioned limiting factors are not present and if sufficient resources are available, a reliable and representative picture of the situation may be obtained from communities randomly selected, for example, from a list of villages or a map. The stay in each village may require two to three days plus travelling time.
Desk research and interviewing professionals. Prior to making the field visits, secondary information should be collected from published or unpublished documents, i.e. papers, reports, theses and newspapers. Geographical and meteorological data as well as maps are the most elementary but fundamental material to be collected. Valuable practical advice can be obtained from experienced professionals who have previously worked in livestock development, disease control, "on farm" research and other community-based programmes.
Discussion of study objectives with authorities. Central or local administrative and technical authorities should be informed about the scope of the study and asked to contribute their comments to better identify the objectives. Dedicated officers can be an excellent source of information on the livestock situation in the study area (including census data, existing legislation and impact of previous government policies) or on the limitations of official statistics. For example, where livestock ownership is taxed, farmers may be reluctant to give information on the composition of their herds and vaccination campaigns or dip tank records may provide only rough estimates about animal populations, and probably only those of large ruminants. Approval and clearance by national or regional offices is indispensable for the success of the activity and in order to benefit from the assistance of district and local authorities. Central authorities should be informed about the areas to be visited for the RA so that they may facilitate the field trips of the team. They should also be informed about the confidential nature of the data to be collected, but at the same time they should be assured that they will benefit from the RA results.
Preliminary visit to the communities. Correct information from farmers can be collected only if their trust is gained. The preliminary visit has as its main purpose the presentation and discussion of the study objectives with the community and local representatives and the collection of background information on the local farming and livestock systems before deeper investigations are begun. Since it may happen that farmers are unwilling to collaborate or that they are too busy at that time of year or that most of the heads of household or animals, i.e. contract herds in West Africa or transhumance, are away, interviews should be arranged when convenient for the farmers, for example, around main festivals. Informal leaders such as religious leaders, chiefs and representatives of social organizations should be also contacted where feasible, since their support is often indispensable to gaining the farmers' trust and because in some areas government authorities may be disliked. Their social role should always be acknowledged and respected as no collaboration is likely to be forthcoming without their assistance, particularly in remote areas.
During this preliminary visit, the main environmental features as well as social and ethnic groups may be identified and recorded and key informants interviewed. Key informants are persons who have specialized knowledge not shared by most of the community (Edgerton and Langness, 1977). Examples of key informants are various formal and informal leaders, representatives of farmers' and women's associations, teachers, veterinary assistants, slaughterhouse inspectors, agricultural and health workers, traditional animal healers, livestock traders, successful and respected individual farmers as well as poor farmers, religious leaders, ethnic and clan leaders and village elders. The main information to be systematically collected should include:
· land forms and the relationship between land form and land uses (for crop production, grazing, etc.);· local cropping calendars and agricultural practices. Livestock production, constraints and husbandry practices should be analysed within the framework of the farming system and related to other agricultural activities;
· seasonal variations in labour demand (for men and women) and main festivities. Meetings with farmers and eventual project activities involving their active contribution should be concentrated in the slack periods of the year. Animals are slaughtered mostly for ritual purposes during festivities and their postmortem inspection may provide information and specimens. Some livestock operations may also be ritually performed on such days; for example, in the central highlands of Ethiopia bulls are castrated on Maskal, an important local festival (Ghirotti and Woudyalew, in press);
· gender division of labour, particularly livestock-related tasks;
· main environmental changes and social events of past years, e.g. historical events, introduction of new farming practices, disease outbreaks that occurred in the area. These may have caused changes in production strategies, husbandry practices, livestock performance or disease occurrence. Such information will assist in the interpretation of data collected in the later phases of RA;
· species and breeds of livestock kept and main husbandry practices, including spatial distribution and changes in past livestock species and breeds. This may have resulted in differences in productivity and land utilization, as well as in the presence or absence of species and breeds, which can be used as indicators, e.g. browsers rather than grazers as a sign of land degradation, and the degree of susceptibility to diseases such as trypanosomiasis, tick-borne diseases, streptothricosis and foot-and-mouth disease;
· purpose for keeping each animal species and their role in farming systems and household economics;
· average productive parameters and their seasonal trends such as fertility, milk or egg production, productive career and different types of offtake (rough percentage 'of sales, exchanges, gifts or slaughter);
· estimation of number of herds and households present in the area and distribution of livestock ownership;
· vernacular names of the most common diseases in livestock and in humans, their importance and spatial/temporal distribution, local health care systems and beliefs;
· main development constraints and public health problems and their spatial/temporal distribution;
· development programmes and facilities present in the area, e.g. crushes, dip tanks, veterinary clinics and weighbridges, and arguments in favour or against existing or eventual development activities;
· food of animal origin most commonly produced, consumed or sold, including existing food taboos and use of other animal products or by-products, e.g. manure, rumen content, horns;
· main sources and availability of feed, including use of by-products, and watering;
· distribution, nature and composition of present wildlife populations. The presence or absence of different wild animal species in the area can be used as an indicator of the degree of pressure placed on the ecosystem by human settlements and activities. Wildlife is often a source of animal products, alternative or additional to the ones from livestock, e.g. bush meat;
· presence and distribution of pests and vectors of diseases.
The existence of a commonly used local calendar should be investigated so that the interviews can refer to such a calendar. Traditional calendars relate to seasonal changes and therefore to agricultural activities. Assisted by these calendars, farmers will remember past events better and thus give more precise answers. For example, local calendars in the Ethiopian highlands begin in August-September, which corresponds to the end of the rainy season when grains are about to be harvested.
Seasonal prices for agricultural products could be collected at local markets (Ghirotti, 1988).
Setting final objectives. From the results obtained from the preliminary visit and the suggestions provided by the contacted community, the final objectives can be identified by the team. Questionnaires can now be designed and further background information on the subjects to be investigated can be collected. Preferably, some days should pass between the preliminary visit and data collection at household level, especially if the team members are not already known to the community. Under some practical circumstances, however, for example, for project formulation missions, this is not possible, and the community should be informed before the visit about the study objectives.
Data collection. Data collection involves the testing of already formulated hypotheses and the identification of new ones. Data can be obtained and checked through direct observation of selected indicators, case histories and interviews with groups of farmers or animal owners, among others. Household visits and interviews can provide a large amount of valuable data using an interview schedule to standardize questions and asking for explanations and opinions from the farmers. Women and perhaps children should be interviewed also since they are often responsible for small ruminant, poultry, pig and rabbit breeding. The interview schedule should concentrate on a few selected quantitative features that, together with the information already gathered through the semi-structured interviews, can give a good picture of the situation. The herd structure and its main parameters can synthesize the health and production status. For ruminants, data to be recorded from each herd/household are the number of:
· calves, kids or lambs (animals below one year of age) that were born and died within the last 12 months;
· adult females of reproductive age (in traditional systems, cattle above four years of age and sheep and goats above one year);
· adult females not of reproductive age (heifers between two and three years of age);
· adult entire males (above one year of age);
· adult castrated males, e.g. oxen;
· adults that died within the last 12 months:
· animals sold, slaughtered or given away within the last 12 months.
The same data should be recorded for equines. If the sample of animals is large enough it is possible to estimate herd fertility rates, i.e. calving, kidding and lambing percentages, mortality percentages below or above one year and offtake rates. In other words, the potential level of expansion or contraction of the herd can be determined. For further discussion on this topic, see Matthewman and Perry (1985) and Baptist (1990). The relative proportion of the different age/sex classes can provide additional information, not only on herd growth, but also on the main reasons for keeping livestock. If possible, the monthly distribution of births and deaths should be obtained as well as the main causes of losses.
For poultry, information should be collected on the number of adults and chicks owned, the number of chicks that were born and died during the year and the number of adults sold, slaughtered or given away. Bee-keeping and breeding of domestic rodents are often other important sources of food and income. In some disease control programmes, e.g. for rabies, pet ownership may be investigated. As mentioned, the interview content can be refined according to the selected study objectives and the earlier results of the field work. For sampling purposes, two different clusters can be chosen for livestock data collection: the household and the grazing unit.
In the former case, the investigation pattern is to first identify the household whose herd might be studied, while in the latter, the herd is first identified before the owner is approached. The former choice is recommended, notably in exploratory RA, when analysing general livestock conditions in a given community or farming system (overall use of animal resources involving different animal species). Groups of households should be selected, preferably stratified according to income indicators (housing conditions, ownership of luxury goods, etc.) or ethnic and cultural features. Also, when land assets are not communal, households should be identified to investigate the livestock situation.
Alternatively, when the analysis is to focus on selected livestock species, e.g. cattle, investigations can be carried out on herds that gather daily in communal pastures (the grazing units). Each grazing unit must be considered as one herd. For example, in the Ghibe Valley and in the Sidamo midlands of Ethiopia, on average, only one herd out of every five includes a mating bull. If each herd is considered separately and not as part of a more complex unit, it is difficult to understand not only the reproductive performance of the single herds, but also the choices of different farmers and the overall dynamics of the livestock system (Ghirotti, 1988). The factors involved in the formation of such grazing units vary and their analysis can assist in understanding some of the social and productive features of the community. Some of the main factors are the ethnic or religious origins of the owners, location of the kraal, age/sex groups (calves may be herded separately near the village and milking cows may be given better pasture) and fodder or herder availability. Where pasture is communal, the steps in herd selection are the following:
· estimate how many grazing units there are in the area;
· identify the reasons for their formation;
· decide which have to be studied based on these criteria;
· analyse their composition;
· make a list of the livestock owners;
· interview a random selection of owners if there are too many herds.
Individual farmers should be interviewed preferably at their homes using questionnaires. During such visits, selected indicators may be checked directly (housing conditions, farm size, herd size and composition), verifying the collected information in a discreet way.
Some external specimens can be collected for further investigation, such as faeces and ectoparasites, and measurements can be taken, including milk offtake using a graduated jug, body scores and weight estimates based on heart girth measurements. Milk can be utilized to perform immunological tests, such as the milk ring test for brucellosis, whose results can give an indication of the herd husbandry conditions. Domenech, Lucet and Courdet (1982) found a high correlation between knee hygroma and abortion caused by brucellosis in cattle. At this stage of RA studies animals should not be approached too closely or insistently as this may irritate the farmers.
A few selected case histories can be recorded on the spot to check replies and obtain more open answers from the farmers about their decision-making processes and production strategies. Examples of questions are: "When was this calf born? Which is its mother? How old is she? How many other calves has she had before? How many of them survived and how many died? Why did they die? What has happened to the ones that survived? How much milk does the cow produce daily for human consumption (milk offtake)? In which month does the cow produce more milk and in which month less?" Another line of questioning may be: "Why is this sheep sick? Since when? What are the symptoms of the disease you have named? Is it frequent in this area? What are the causes? When and where does it occur? Do other animals and species suffer from the same disease? Which are the most vulnerable ones? How many of the affected animals died? Is it a major livestock problem? Which other diseases are more important and which are more frequent in this area? Why are the diseases seasonal in occurrence? Which animals are affected?"
Because of the even greater sensitivity of pastoral societies to questions about livestock ownership, interviews with nomadic and transhumant people should not focus on herd or flock sizes (information that can be otherwise obtained through periodical direct observations, i.e. aerial surveys, counting and checking at watering, vaccination or dipping points), but rather on production dynamics (seasonal distribution of events, such as calving or mortality, and main reasons for offtake and culling) and qualitative information, including epidemiology of diseases, husbandry practices and the nature of factors limiting livestock exploitation.
Data analysis and interpretation of results. The data obtained from the interviews should be analysed as soon as possible and the results compared with the information obtained through the semi-structured interviews. Unexplained differences should be investigated, and any hypotheses should be made and tested, and discussions held with farmers, while the team is still on the site. It should be emphasized that information gathering and analysis is a continuous process that takes place at every step of RA. Arithmetic means and rates can be easily calculated and distributions tabulated using pocket calculators. Moreover, modern computer technology allows the utilization of laptops in which spreadsheet software can be run. Herd projections can be carried out through rough "what if?" analysis, making use of the different productive parameters recorded. Comparisons between grazing units and households can be made by converting different animal sizes and species into tropical livestock units (TLU). A TLU is commonly taken to be an animal of 250 kg live weight (see Table).
Feedback and discussion with the community. The results of the appraisal should be openly discussed in summing-up meetings with the farmers concerned, bearing in mind the potential target groups of the proposed supportive measures and different types of bias associated with such forms of discussion. Through the continuous hypothesis-verification process, the team should ensure that the picture of the livestock situation it obtained in a given area is not a misleading one, and seek confirmation and explanation of recorded data from the community.
Discussion of the results with central authorities and colleagues. The answers gained through RA should lead to the identification and selection of practical activities already discussed with the concerned communities and local authorities, however, such solutions should be further discussed with the proper government authorities and potential donors. Little can be achieved at community level without the support of informal leaders, and little can be achieved at central level without the support of government authorities and sometimes that of academics.
Dissemination of collected information. Information can be shared and results discussed in reports, workshops and meetings with colleagues or other officers involved in applied research and development programmes. Through these avenues, it is possible to receive useful criticisms and suggestions both for further applied investigation and for programme formulation. At this point, a project document may be formulated.
Research or development project phase. As mentioned earlier, RA should aim at identifying practical activities to improve living conditions of the communities concerned. If responsible authorities approve them and resources are available, some activities identified by the team with the assistance of the community may be implemented. In such cases, the results of the earlier RA can provide a basis for further applied and better targeted studies and for the establishment of monitoring systems to assess the progress of planned activities.
Follow-up. If developmental or research activities are to be carried out, they should be monitored in time, e.g. during the implementation of the project, together with the parameters of herd dynamics, so that a more accurate picture of the changing situation can be drawn up with the farmers and activities corrected as needed. The RA should have shown the participation and confidence levels of local farmers toward researchers. If farmers become increasingly more confident, more accurate sampling and measurements can be gained, including those through direct contact with farmers' animals. In such case, blood samples can be taken for different immunological, microbiological or biochemical tests. The farmers who have proved to be collaborative can be involved in the study or in pilot project activities, and reasons for lack of interest from others can be investigated. In this way, RA gradually becomes a tool to understand the community and then, with its assistance, to carry out activities that meet expressed needs and to evaluate results.
Several types of diagrams are widely used to summarize and present the collected information so that it can be discussed, often even in the field. Besides histograms and bar and pie charts, the ones most used are:
· transects, where the most relevant features of the different areas and ecological zones analysed are summarized. These are particularly useful in showing spatial differences and trends (Figure 2);· maps, showing seasonal movements of animals, water points, livestock facilities, grazing areas and permanent settlements;
· seasonal calendars, where the occurrence of and changes in human activities, production and biological events, including diseases, are plotted against climatic data. These can be useful in highlighting temporal patterns of phenomena (Figure 3);
· flow diagrams and decision trees, where the key factors influencing decision-making and the consequences derived from such decisions or other changes can be shown.
The investigation of livestock is the investigation of one of the most sensitive subjects for farmers because of the fundamental socio-economic and cultural role of animal resources in most traditional societies. Questions about the size of their herds or flocks are very likely to raise suspicions about future taxation or other unpleasant consequences: the analogy made between a herd and a bank account in the industrialized world is a common one. The endorsement of the team by the community, therefore, is the crucial facor when performing RA in animal health and production. Community leaders and key informants can play a critical role as mediators, guaranteeing the confidential nature of RA while explaining eventual benefits.
Average tropical livestock unit conversion factors for different species - Facteurs moyens de conversion en unités de bétail tropical pour différentes espèces - Factores medios de conversión por unidad ganadera tropical para las diferentes especies
|
Species |
Tropical livestock unit conversion factor |
|
Cattle* |
0 7 |
|
Sheep or goats |
0.1 |
|
Horses |
0.8 |
|
Mules |
0 7 |
|
Donkeys |
0.5 |
|
Pigs |
0.2 |
|
Chickens |
0.01 |
*Calves = 0. 15.
As previously discussed, some of the main methods of obtaining primary data through RA are: direct observation of selected indicators; semi-structured interviews with key informants; individual interviews, mostly through questionnaires on selected items; group interviews; and workshops.
Errors associated with the use of questionnaires and direct observations have been analysed by Thrusfield (1986) and Putt et al. (1987). Bias can be reduced through careful and correct sampling and questioning. Questions about livestock numbers should be kept until the end of the interview. Because of the inevitable distortion arising from an interview with a limited number of key informants, the information so collected should be checked and investigated using other field techniques. The presence of outsiders influences the behaviour of local people and their responses, which they may alter to please, confuse or deceive the researchers; this phenomenon is known as the Heisenberg uncertainty principle. Especially during group interviewing, expectations may increase and answers may reflect what people wish more than what they actually know and think. Literate people and élites may receive more attention in meetings than the common farmers (Edgerton and Langness, 1977), and in some societies women are not allowed to speak freely. On the other hand, people in some communities may live in fear and may view outsiders as government officials, therefore, they will avoid expressing their opinions openly during public meetings or they may give wrong information. Mitchell and Slim (1991) argue that in developing countries interviewing involves two kinds of structural bias derived from Western culture: expecting answers and "nutshelling". The first is the assumption that each question has a very straightforward spoken answer, while the second concerns the concept of brevity: we are accustomed to short questions and short answers. In many traditional societies brevity is not accepted, and communication means discussing main events and phenomena rather than responding to a number of brief questions with brief answers.
Because of their nature - mostly qualitative and using purposive sampling - rapid assessment methodologies are not a substitute for standard investigation techniques, but they can be valid complements. RA should be performed before any large-scale baseline survey is begun, and its methodologies should be adopted in formulating and monitoring development activities.
Veterinary science has always benefited from the contribution of other disciplines. Its main scope is the study and control of factors jeopardizing better utilization of animal resources. In recent years greater emphasis has been put on the management of available means, especially human resources, in disease control actions (Bainbridge and Sapirie, 1974; WHO, 1990). Social sciences such as anthropology and sociology have provided useful methods and they have much in common with epidemiology, the latter being an indispensable tool both culturally and operationally for every veterinarian. Social scientists, like epidemiologists, are concerned with patterns. The former investigate cultural and organizational patterns within a community, while the latter investigate patterns of disease and other limiting factors within a population. Both deal with large numbers or groups and build their knowledge on field work. In development, rat-her than recording definite average values, they are more interested in comparing and identifying determinants that influence patterns and differences. Properly designed and investigated case-studies, in which different factors are systematically analysed, may provide operative answers to practical problems. Today, more often than in the past, animal health and production officers are asked to work closely with other professionals in development activities and to-provide practical solutions to straightforward problems. Field methods that are cognizant of the farmers' viewpoints and that encourage collaborative work with the concerned community assist in carrying out such a demanding task.
Abramson, J.H. 1984. Survey methods in community medicine, 3rd ed. Edinburgh, Scotland, Churchill Livingstone.
Bainbridge, J. & Sapirie, S.1974. Health project management. Geneva, World Health Organization.
Baptist, R. 1990. Simulated livestock dynamics - effects of pastoral offtake practices and drift on cattle wealth. Trop. Anim. Health Prod., 22: 67-76.
Carruthers, I. & Chambers, R.1981. Rapid appraisal for rural development. Agric. Admin., 8: 407-422.
Domenech, J., Lucet, P. & Courdet, M. 1982. La brucellose bovine en Afrique centrale: V. Description d'une méthode d'enquête simplifiée Rev. Elev. Méd. vét. Pays trop., 35: 125129.
Edgerton, R.B. & Langness, L.L. 1977. Methods and styles in the study of culture. San Francisco California USA, Chandler and Sharp Publ.
Ghirotti, M. 1988. Farming systems, household economics and child malnutrition in Sidama: an international course for primary health care managers at district level in developing countries. Rome, Istituto Superiore di Sanità.
Ghirotti, M., Semproni, G., De Meneghi, D., Mungaba, F.N., Nannini, D., Calzetta, G. & Paganico, G. 1991. Sero prevalences of selected cattle diseases in the Kafue flats of Zambia. Vet. Res. Comm., 15: 25-36.
Ghirotti, M. & Woudyalew, M. "Madosha" - traditional castration of bulls in Ethiopia. In E. Mathias-Mundy, C.M. McCorkle & T. Schillhorn van Veen, eds. Ethnoveterinary research and development. London, UK, Kegan Paul Int. (In press)
ILCA. 1977. East African. pastoralism - anthropological perspectives and development needs. Addis Ababa, Ethiopia, International Livestock Centre for Africa.
Jahnke, H.E. 1982. Livestock production system and livestock development in tropical Africa. Kiel, Germany, Kieler Wissenschaftsverlag Vauk.
Matthewman, R.W. & Perry, B.D. 1985. Measuring the benefits of disease control: relationship between herd structure, productivity and health. Trop. Anim. Health Prod., 17: 39-51.
McCracken, J.A., Pretty, J.N. & Conway, G.R. 1988. An introduction to rapid rural appraisal for agricultural development. London, UK, International Institute for Environment and Development.
Mitchell, J. & Slim, H. 1991. The bias of interviewing. RRA Notes, 11: 20-22.
Putt, S.N.H., Shaw, A.P.M., Woods, A.J., Tyler, L. & James, A.D. 1987. Veterinary epidemiology and economics in Africa. ILCA manual no. 3. Addis Ababa, Ethiopia, International Livestock Centre for Africa.
Schwabe, C.W. 1978. Cattle, priests and progress in medicine. Minneapolis, Minnesota, USA, University of Minnesota Press.
Thrusfield, M. 1986. Veterinary epidemiology. London, UK, Butterworths.
WHO. 1990. Health systems research and intersectoral collaboration in veterinary public health. Report of a WHO Consultation, Geneva, 4-8 December 1989. Geneva, World Health Organization.
