Cover Image
close this bookGuide for the conduct of the constraints analysis component. (1997)
View the document(introduction...)
Open this folder and view contentsParticipatory survey methods for gathering information


Annex II
SPFS/DOC/18 handbook series volume II
Special programme for food security -


Informal participatory survey methods can be used for gathering of data at various stages of programme implementation, and for analysis of SPFS-related issues by the farming communities themselves.

The main principles of informal data-gathering include multidisciplinary team work, triangulation and flexibility.

Multidisciplinary team work is carried out by a team composed of individuals with different professional backgrounds. This team is in charge of collecting and analysing data from rural environments.

Triangulation is the intentional collecting of information from several different perspectives. This can be achieved through including in the multidisciplinary team individuals with different levels of experience and gender; selection of different units of analysis (e.g. farmers' groups, households, or individuals) and use of different techniques (e.g. scoring, mapping, diagramming).

Flexibility is the absence of a rigid protocol and the possibility to change techniques and tools as needs arise. At the same time, it must be remembered that rapid and participatory data collection exercises have to be well-planned.

In addition to these core principles, serendipity and investigation of exceptions are also often used in data gathering. Serendipity is the creation of situations in which the researcher can take advantage of chance, while investigation of exceptions means following up on exceptions and lack of agreement.

Preparations for any informal participatory survey should cover the following points.

· Decide on the objectives of the survey and the role of the survey team(s) before beginning

Without clear objectives there is a danger that the exercise will be unfocused and lead to feeble conclusions. If the survey team is only gathering information, it will have different objectives and require different composition than a team that is also facilitating farmer assessment of the results and impact of the demonstrations.

· Establish the survey team(s) on the basis of the defined objectives

Factors to be considered include team selection (right mix of experience, gender, and discipline), team training required, team size (not more than two or three), number of survey teams carrying out surveys in parallel (not more than three), and procedures for note taking and report writing.

· Consider stratification of the survey area or population if not all locations, individuals, or groups in the demonstration areas are to be included in the survey

Stratification involves dividing the survey area or population into subsets within which the variability of key factors is expected to be lower. A sample of individuals, groups or locations for actual survey can then be selected from each subset. A typical example is that of agro-ecological zones where an area is divided into subsets based on homogeneity of rainfall, fanning systems or even crop yields. At the same time it is important to avoid biases, such as road conditions, when choosing locations.

· Decide in advance whether to have farmers' groups or individuals as the unit of investigation

This will depend upon the circumstances and the subject of analysis. Group discussions are particularly useful if the information required is not expected to vary widely between households. If the issue under discussion is not too sensitive in nature, a group interview may provide an easy overview of likely variables, such as field sizes or crop yields.

· Make use of results from previous work (secondary data)

Secondary data can improve the efficiency of surveys by raising their explanatory value and avoiding the effort of gathering the same information twice. In order to avoid being misled by rumour, myth or gossip, it is essential to support and cross-check findings by direct observation of important indicators.

Two major categories of informal participatory appraisal methods that are particularly useful for gathering general socio-economic information about a demonstration area in the early stages of the pilot phase are semi-structured interviewing with checklists, and mapping and diagramming. Two more focused methods involve surveys of farmers' knowledge, attitudes and practices and stakeholders' analysis. Finally, quantitative survey methods are sometimes needed to complement the more informal participatory methods.

Semi-structured interviewing with checklists

Semi-structured interviewing is a form of guided interviewing in which only some of the questions are predetermined and new questions are usually generated during the interview. Interviewers use a checklist of questions as a flexible guide rather than a formal questionnaire. As a result, the interviews tend to take the form of discussions, during which both interviewer and interviewee learn from each other. It is important that farmers participate on an equal footing during the information-gathering process, otherwise their participation during the later stages of the demonstration phase is likely to be undermined.

0 Steps to Follow

The following steps should be followed in semi-structured interviewing:

Before the survey:

· select the multi-disciplinary survey team
· analyse secondary data
· prepare checklist for the interview (this should be a team exercise)
· prepare the logistical side of the survey
· inform farmers in advance
· establish note-taking procedures before entering a village
· decide whether group discussions or individual in-depth interviews are more appropriate

During a group meeting or individual interview:

· be aware of the local culture and language

· respect farmers as equal partners - listen to them and what they have to say

· do not use the checklist as a questionnaire but as a means to stimulate discussion and a participatory dialogue

· build the questions to be asked around a list of sub-topics, existing information on the community, or visual material such as diagrams, photographs, or maps

· use the "six little helpers" for probing: who? why? what? when? where? how?

· take notes during the interview but not excessively (sometimes it is better to complete notes immediately after the discussion)

After the meeting or interview:

· finish the discussion politely

· have evening brain-storming sessions with the team to complete notes and to prepare for the following day

· establish report writing procedures with the team making sure enough time is allocated

1 Checklists

Box 11.1 is an example of a checklist used for semi-structured interviewing. Each survey team should establish its own checklists depending upon what data the team considers necessary. Checklists should not be long and unwieldy. It is often better to concentrate on a number of key issues only, but cover these well.

Box II.1

Example of a checklist to collect information on farming system in demonstration area


Ö Location of demonstration area
Ö Agro-ecological zone
Ö Main economic activities of local population


Ö Physical resource base: land, soil, water, vegetation, etc.
Ö Land use patterns: agriculture, livestock, forestry activities
Ö Cropping patterns: crops, varieties, patterns, rotations, varietal preferences
Ö Assets available (e.g., major tools)
Ö Yields per crop per unit of land
Ö Quantities of physical, variable inputs used per crop per unit of land
Ö Labour used per crop per unit of land
Ö Prices for: inputs, outputs, labour, land, capital


Ö Distribution of assets
Ö Gender roles
Ö Land tenure system
Ö Interest groups
Ö Labour use patterns
Ö Access to services and markets

Ö Cultural attitudes towards farming

Aside from their use in semi-structured discussions, checklists are also a part of other participatory tools, such as scoring/ranking and mapping/diagramming.

A useful reference for further information on semi-structured interviewing is: Natural Resources Institute (NRI), Participatory Rural Appraisal - A Manual on Issues, Principles and Tools (1996), NRI, Chatham, UK.

Mapping and diagramming

A diagram is any simple schematic device which presents information in a condensed and readily understandable visual form. It is a simplified model of reality.

Diagrams summarize data in a way that can be used at all stages of the SPFS. They are most useful as an aid to participatory discussions with farmers and other community members in demonstrations. Diagrams can be drawn on almost anything - for example, on paper, overhead transparencies, blackboards or sand, depending on the situation.

It is important to draw the diagrams in the presence of different categories of people (women and men, young and old, etc.), as their perceptions, viewpoints and information will often differ. Comparison of the diagrams drawn by different people can lead to a deeper understanding of the diversity of opinion and decision-making processes in a community.

Diagrams are useful tools because they:

· greatly simplify complex information
· facilitate communication
· stimulate discussion
· increase consensus among team members
· are an excellent way of involving community members, and determining their views and categories, especially if they are encouraged to draw diagrams on their own

Box II.2 lists possible types of diagrams.

Box II.2
Possible diagrams




map, transect


seasonal calendar, daily activity calendar, time trends, historical profile


flow diagram, sources of livelihood chart, farming systems diagram


problem tree


decision tree, Venn diagram

2 Maps

Maps are diagrams showing the geographical arrangement of key features of an area of land. They are a useful way to gain understanding of systems at different levels: individual fields, farms, communities or districts. When villagers draw maps they include the features that they feel are most important. For this reason, different stakeholder groups will often draw maps of the same area that look very different. As a result, mapping can be a good way to gain insight into the priorities of stakeholders, conflicts of interest, and possible opportunities.

There are many equally valid ways to produce a map with a group of villagers. Before beginning the exercise it is important to draw up a checklist of the major features that the map should include. Common features of maps are:

· physical infrastructure
· social infrastructure
· cropping systems
· water sources
· woodlands
· major physical features
· land tenure systems
· grazing areas

Participants, location and equipment should all be considered:

· Participants: It can be useful to do the exercise with separate groups of the village in order to obtain an unbiased view of the subject. Groups should generally not be too large (no more than 15) for this exercise.

· Location: Mapping exercises often take quite a long time and it is important to conduct them in a comfortable place. The location should also be reasonably peaceful and free from distractions.

· Equipment: Maps can be prepared with a variety of tools on a variety of surfaces. The most common combinations are:

- sticks, stones, leaves, etc. on cleared smooth areas of ground
- coloured chalk on cement floors
- coloured pens on paper

Each has its advantages and disadvantages. For example, ground doesn't have edges but it may be rained on, chalk is easy to rub out and re-draw but there may not be a cement floor available, paper is easy to keep after the map is finished but making changes on paper is difficult. The best choice will depend on the particular circumstances in a village.

In mapping exercises, the following steps are important:

· the purpose of the map must be carefully explained to the participants

· once the map is started, guidance should be kept to the minimum although it may be necessary to give some assistance with the first features

· the map is a tool and not simply a product; as the map emerges, discussion can take place on its features and questions can be raised about items on the checklist

· if the map hasn't been prepared on paper, it will have to be copied when it is completed; this should be done very carefully and the copy should be as exact as possible

· a copy should be shown to the participants and a copy should be left with them; it is their map and experience shows that participants prefer to keep their work

Boxes II.3 to II.5 give examples of different kinds of maps which may be useful for the SPFS.

For more detailed information on drawing different kinds of maps, see Bonnal, J. and M. Rossi (principal authors). Guidelines for Participatory Constraints Analysis at Community and Farm Household Level (SPFS /DOC /17, preliminary working draft, 1996), FAO, Rome, Wilde, V. and V. Mattilla, Gender Analysis and Forestry International Training Package (1995), FAO, Rome, and Nabasa, J., G. Rutwara, F. Walker and C. Were, Participatory Rural Appraisal: Practical Experiences, Natural Resources Institute (NRI), Chatham, UK.

Box II.3 Village Map (example from Butawata Village, Mubende District, Uganda)

a) Drawn on ground by community members

b) Formal copy identifying different households, water sources, places of worship and shops

BOX II.4 - Village social map, wealth ranking (example from Ethiopia)

Box II.5 - Village resources map (example from Khajret - Uperli Guanguri, Himachal Praelesh, India)

0 Transects

A transect is a diagram that is produced during a walking discussion with villagers and shows the key features of different land use zones in a community. Transects are particularly useful when there is a range of land use systems in one community. This is often the case when communities are located on the coast, in hilly areas, on rivers or lakes, or in areas where soils vary over short distances.

Transect walks are a good way to:

· analyse demonstration areas and their suitability for water development improvements and changes

· gain a basic understanding of the agriculture in an area and its constraints and opportunities

· establish contact with villagers and engage in a semi-structured discussion with a villager or group of villagers

Transect walks offer an opportunity for direct observation by participants. It is very important to keep one's eyes open during the walk.

The first step in preparing for a transect walk is to draw up a check list of the areas of main interest for the exercise. Common areas for transect check lists are:

· soils/fertility
· slopes
· food crops
· cash crops
· trees
· water availability
· land tenure
· livestock
· problems
· opportunities

The route should be planned with the villagers. It does not have to be a straight line, but it should not be random. It should be chosen to pass through the main land use systems. It may be best to use a route that brings the team back to where it started, or it may need to start beside a river or the sea and move up-hill to the edge of a watershed. Sometimes several short transects give a better overall picture than one long one.

Recording is quite difficult during a transect because the team members are walking and because the whole check list needs to be covered each time a new land use system is traversed. Stopping on the way when interesting issues arise is important. It is also useful to divide up responsibilities among the team members. For example, one may take responsibility to ask about crops, another about land tenure and so on.

The diagram should be prepared as soon as possible after the walk is completed. The rough topography of the walk should be drawn at the top and a matrix with the headings of the check list laid out underneath. If possible, villagers should be involved in the preparation of the diagram. In any case, it is important to show the diagram to groups of villagers and use it as a basis of discussion on what the team has learned. It will help focus the discussion and enable team members to probe further on the fanning systems, the severity of constraints and the degree of consensus amongst villagers.

Box II.6 gives an example of a transect.

Box II.6 - Example of a Transect

More information on transects can be found in: Nabasa, et al. Participatory Rural Appraisal: Practical Experiences, op. cit.

1 Seasonal and Daily Activity Calendars

Calendars are diagrams showing the timing and/or importance of events over a period of time (a year, a production season, a day). One diagram can include several different variables.

There are many types of calendars and many uses. They are an excellent way of understanding and discovering variations in behaviour through time and their relationship to other events.

Seasonal analysis is particularly useful for understanding constraints and major problems that might emerge only at certain times of the year.

A checklist should be prepared of the variables to be included in the calendar and any associated issues to be discussed about each of them. Some examples of variables which might be "relevant for a seasonal calendar are:

· food availability
· rainfall and temperature
· production and post-harvest activities
· prices
· marketing
· income
· labour or levels of work activity
· debt

Participants, location and equipment should all be considered.

· Participants: seasonal calendars can be prepared by individuals or small groups. It is important to identify participants who represent the group you wish to consult.

· Location: the preparation of seasonal calendars does not need a lot of space. They should be prepared in the shade in a comfortable spot.

· Equipment: seasonal calendars can be drawn on paper, but it is usually best to draw them on the ground with a stick or chalk and to use locally available materials such as sticks, stones, beans, leaves or fruits to represent the different variables.

In preparing seasonal calendars, the following steps are important:

· the purpose of the exercise should be carefully explained to the participants

· the horizontal axis, representing time and divided into the different seasons, should be drawn first (In most rural societies, the seasons are not thought of as calendar months, but have their own names. These should be used for the diagram. If seasons differ in length, then they should occupy roughly proportionate lengths on the horizontal axis.)

· the calendar should be drawn for more than one year, to ensure that the cyclical nature of events is correctly understood (Calendar months can be added after the calendar has been completed or after the horizontal axis is agreed.)

· the variables to be included should be placed on the calendar one by one (Some variables, such as planting or weeding, can be represented simply by showing the times when they occur. Other variables, such as rainfall or prices, require quantitative estimates to be shown as well.)

· a dot or horizontal line should be used to mark the relevant time period on the calendar

· the amount of time can be displayed using stones, sticks or seeds or a vertical axis to indicate quantity

The seasonal calendar is a tool and not simply a product. Once it is completed it should be used as the basis for discussion about the variables included within it. Seasonal calendars are easy to record. Notes on the discussions of the calendars should also be made.

Box II.7 gives an example of a seasonal calendar.

A daily calendar is similar to a seasonal calendar except that it shows the pattern of daily activities of community members at different times of the year. It also enables a comparison of the daily routines and work loads for different groups of people (e.g. women, men, young, old, employed, unemployed, etc.) It helps to identify time constraints (shortages) and opportunities. For example, it can be used to determine the most appropriate time of day for a women's training course.

Daily activities are often illustrated on daily activity clocks. Box 11.8 gives an example of a daily activity clock at two different times of years.

Further information on seasonal and daily calendars and daily activity clocks can be found in: Theis, J. and H. Grady, Participatory Rapid Appraisal for Community Development (1991), International Institute for Environment and Development (IIED)/Save the Children Federation, London, UK, and Townsley, Training of Rapid Appraisal Teams. Notes for Trainers, (1993), FAO, Rome.

Box II.7 - Seasonal Calendar (example from North Kordofan, Sudan)

Box II.8 - Daily Activity Clocks (example of seasonal daily activities of men and women in Dzinavene, Chivi District, Zimbabwe) - Women - Winter Dry Season

Box II.8 - Daily Activity Clocks (example of seasonal daily activities of men and women in Dzinavene, Chivi District, Zimbabwe) - Women - Summer Wet Season

Box II.8 - Daily Activity Clocks (example of seasonal daily activities of men and women in Dzinavene, Chivi District, Zimbabwe) - Men - Winter Dry Season

Box II.8 - Daily Activity Clocks (example of seasonal daily activities of men and women in Dzinavene, Chivi District, Zimbabwe) - Men - Summer Wet Season

Knowledge, attitude and practice surveys

Knowledge, attitude and practice (KAP) surveys are used to understand and assess farmers' local indigenous knowledge, values and belief systems and how these affect their farming practices. KAP surveys are useful tools for identifying the technological interventions which are important in an area and which are likely to create a significant impact. By analysing fee words farmers use to communicate their knowledge, attitudes and practices in regard to specific elements of a farming system, it is possible to identify those elements which may be good, those which may need to be improved, or those which may need to be discouraged.

Wife this information, demonstrations can be more effectively designed. It may not be necessary, for example, to provide all farmers wife a complete set of technology recommendations, as some of fee farmers may already know, agree wife, and/or have acted upon the necessary information. On the other hand, KAP surveys can identify elements of fee technology package feat are not known to the majority of farmers, fee reasons for their negative attitudes, how and why they have practised recommended technologies inappropriately, and so forth.

KAP information can be obtained from surveys carried out for feat purpose, or can be extrapolated from information already collected using other methods such as semi-structured interviews. If KAP surveys are used for gathering baseline information, fee same survey method should also be followed for monitoring farmers' reactions to pilot phase demonstrations.

In order to carry out a KAP survey, it is necessary to:

· collect information on farmers' knowledge, attitudes and practices regarding their farming systems

· identify fee main topics of importance to fee farmers by counting fee number of times each topic was cited in discussions and interviews

· prepare a topic card for each of fee main topics identified, wife a summary of fee farmers' information about their knowledge, attitudes and practices, noting also any additional information to be collected

Boxes II-9, II-10 and II-11 illustrate in more detail fee steps required to carry out a KAP survey.

After each round of survey, results are discussed wife fee farmers and possibilities for using innovative technologies or farming practices to solve identified constraints are talked about. The farmers' groups then decide what kind of extension messages they would like to receive, and fee form in which they would find these most useful. Using this information, strategic extension campaigns to promote widespread adoption of new technologies and practices of most interest to fee farmers can be planned.

Box II-12 shows how fee KAP survey method is used to plan, implement and monitor a strategic extension campaign.

Further information on use of KAP surveys can be found in Adhikarya, R., Strategic Extension Campaign: A Participatory-Oriented Method of Agricultural Extension (1994), FAO, Rome, Jallade, J., Knowledge, Attitudes and Practices -KAP (draft, 1994), FAO, Rome, and Contado, T.E., Applications of KAP in Improving Extension Impact (1996) Paper prepared for fee Second Informal Consulttation on International Support to Agricultural Extension Systems in Africa, FAO, Rome.

Box II.9 Examples of farmers' comments related to their knowledge

· My land is not as fertile as it used to be

· During the last five years I have noticed a decrease in land fertility

· Decrease in production has been mainly in sweet potatoes

· I do not know what the cause is of this fall in production

· My land is subject to erosion

· I did not know that trees existed that could, at the same time, fertilize the land and supply firewood and fodder for the animals

· These are the trees that are used to fight erosion

· Modem medicines are no good for the animals

· I don't know why my animals are sick

· I do not know where to sell my vegetables

Examples of farmers' comments related to their attitudes

· Associated crops are more secure
· I do not allow my land to fallow because I have too little of it
· I am not interested in making compost
· I do not use organic swine manure as it bums my feet
· The burning of green manure is of no use
· I would like to apply other fertilizing techniques
· I am willing to adopt other technologies to fight erosion
· I am willing to plant trees
· I do not want to buy improved seeds produced by other fanners because I do not trust them
· I would like to increase the number of my animals, especially pigs and poultry
· I do not exchange my breeding animals as they are my family's welfare
· I would like to obtain higher production from my animals

Examples of farmers' comments related to their practices

· I plant sorghum by itself and small millet together with cowpeas
· I plant on the same day all the plants that are grown together
· I let fifteen days elapse between ploughing and sowing
· I do not practise crop rotation
· I let my land fallow from one to two years
· To allow my land to fallow means to abandon it
· I breed goats, pigs and poultry
· I use manure on my land but I do not treat it
· I do not use organic manure
· I utilize ashes from my land
· I spread ashes on my land before the beginning of the rainy season
· I fight erosion by digging holes and pouring water in them
· I get the necessary supply of seeds from my neighbours
· I keep my seed provisions stored in my home
· I feed my animals while grazing them in the country side
· I never give my animals any medicine

Box II.10 Example of main topics of concern identified by farmers







seed conversion


poultry breeding


phytosanitary care






fighting insect pests


inorganic manure


The main topics of concern to the farmers surveyed are, by priority: erosion, firewood and reforestation, seed conservation, and the fight against insect pests.

Box II.11 - Knowledge, Attitude and Practice (KAP) survery topic card

For each topic of concern, please indicate:

Summary of farmers' knowledge

Summary of farmers' attitudes

Summary of farmers' practices

Additional information to be collected:

Box II.12 KAP survey as a tool for linking research, extension and communication with farmers


· Identify farmers' problems and assess their needs for new technology and information

· Analyse results of farmers' priorities regarding technology and information

· Develop/validate technology recommendations and determine Critical Information and Technology Acquisition Package (CITAP)

· Develop data collection instruments for Knowledge, Attitude and Practice (KAP) survey of farmers based on CITAP

· Pre-test KAP survey instruments

· Conduct KAP survey and Focus Group Interviews (FGI)

· Analyse farmers "voting" (i.e. survey) results regarding their knowledge, attitudes and practices on:

- what and how much they know and don't know,

- what they like/believe or don't like/believe and why

- whether or not they properly practice the given technology recommendations, and if not, why they improperly practice or don't adapt them

· Modify and improve technology solutions based on KAP survey and FGI results

· Develop Strategic Extension Campaign (SEC) objectives, messages, and multimedia support and training materials, based on KAP survey and FGI results

· Pre-test SEC Messages, Multimedia Support & Training (MMST) materials

· Analyse farmers "voting" results on the clarity, acceptability and perceived relevance of SEC messages and MMST materials

· Involve farmer leaders in planning and field implementation of SEC activities

· Carry out the Strategic Extension Campaign

Stakeholder analysis

Many projects and programmes have failed in the past because they have not taken into account the different interests of groups in the area in which a programme is being implemented. Members of various interest groups can use their power and influence to prevent objectives from being achieved if they feel that their interests are threatened. An understanding of the interests of different groups and incorporation of this knowledge into the planning and implementation of activities can permit more effective design of programmes and increase the chances of success.

In the SPFS, stakeholders can be defined as groups of people with common interests and concerns with respect to the production, marketing, or consumption of food staples.

Farmers are the programme's primary stakeholders because they are the main focus of the activities and they are the ones who are ultimately affected. However, among farmers, there are important sub-groupings to consider. In the context of food security, the differences between food-deficit and food-surplus farming households are particularly important. These sub-groups are likely to have different priority constraints. Other important sub-groupings with different priorities are those based on gender and those based on assets.

Box II-13 gives examples of common key differences among farm households.

Box II.13
Some examples of common key differences among farm households

buy more food than sell

sell more food than buy

women headed

male headed

also work for others

employ workers

no access to irrigation


rent or borrow land

secure access to land

can not afford external inputs

use external inputs regularly

can not obtain credit

credit available

A complete stakeholder analysis will consider all actors in the commodity chain, from pre-production preparation through production to post-production transformation, storage and circulation, and finally to consumption. Besides farmers, these actors include various private-sector enterprises, public services and institutions and collective and non-governmental organisations (NGOs).

Providers of services in the private sector (such as marketing, milling, de-hulling, de-husking, food and beer manufacture, transport and input provision) will often have a key role in the removal of constraints and the creation of new opportunities for increased farm production. At the same time, they, like other governmental and non-governmental stakeholders, will face both gains and losses from the changes involved in successful introduction of new technologies and other innovations into existing farming systems.

Stakeholder analysis generally follows these steps:

· identification and listing of all primary and secondary stakeholders
· classification of the stakeholder groups through assessment of

- their interests in relation to the SPFS
- the likely impact of the programme on their interests
- their influence or power to affect the programme

· preparation of a stakeholder table based on this information

2 Identification of Stakeholder Groups

The first step in stakeholder analysis is to identify the various interest groups in a demonstration area. It is important to have a good understanding of what is going on and the roles played by the different actors in the communities. Semi-structured interviewing can be useful for identifying the stakeholders in a given area. It is important to bear in mind that a stakeholder group might consist of only one individual and that individuals can belong to more than one stakeholder group.

3 Identification of interests and Concerns

The next step in stakeholder analysis is to identify the interests and concerns of each stakeholder or stakeholder group. Workshops involving as many potential stakeholders as possible are useful for sharing information and clarifying of different points of view. On the other hand, stakeholder analysis often involves sensitive information, and as a result many interests are covert and agendas at least partially hidden. Few benefits are likely to be gained by trying to uncover these agendas in public. In this case, it is better to interview stakeholders individually.

The following issues should be considered in doing the analysis:

· The interests of any stakeholder group in the programme depend on the effects its members feel the programme will have on them. The anticipated effects may be positive or negative. Effects may be material (increased income or food availability, lost trade, higher costs) or social (improved or lost status or power).

· Care must be taken in listing and assessing the interests of each stakeholder group. The group's perceptions of the programme may not be the same as those of the analyst. Stakeholders may not willingly express or talk about possible negative effects that are of concern to them. It is important to think about how the activities of each group relate to the objective of improved food security and to the problems that are preventing its attainment. An important reason for conducting stakeholder analysis is to enable both the programme managers and the stakeholders themselves to recognise and better deal with potential conflicts of interest.

· It is crucial to assess the likely impact of the programme on the stakeholder groups. For example, some groups may not be able to adopt a new technology because they do not have access to the resources needed to use it. These groups cannot then gain from the programme, even though it may increase food production in the community overall.

· Finally, the power and influence of each group must be recognized and assessed. Within any community the most powerful and influential can help make things happen and can also prevent them from happening. Conversely, groups with little influence will find it much more difficult to achieve their goals if these involve concessions or contributions from others.

Box II.14 - Format for a Stakeholder's Table for the SPFS

Stakeholder's group


Power and influence in the

Potential positive impact of SPFS intervention

Potential negative impact of SPFS intervention

Preparation of a Stakeholder Table, Linkage Map or Conflict and Partnership Matrix

The information on the stakeholder groups can be summarized in a table which lists the various stakeholders, their interests, their power and influence in the community and the possible positive and negative effects of the SPFS on them. Box II-14 presents a format for a stakeholder table that has been designed specifically for the SPFS.

Drawing a map of the linkages between a peasant farmer and all other stakeholders in the production chain can help to illustrate graphically the various interrelated interests of the different actors. Box 11.15 gives an example of a communication linkage map.

This information will help in evaluating whether a particular demonstration has a good chance of succeeding, or whether it needs modification. It will also highlight potential problems against which it will then be possible to guard.

Another useful exercise is the development of a stakeholder conflict and partnership matrix, which helps demonstrate the areas of conflict and partnership among the various stakeholders. The matrix also helps show whether the extent of the conflict is large or small. Although it is impossible to avoid conflicting interests entirely, this information can be used to help avoid or mitigate situations of severe conflict which could negatively affect the programme. On the positive side, it can serve to help promote partnerships among the various stakeholders. An example of a Venn diagram for stakeholder analysis in shown in Box II.16, and, in Box II.17, of a stakeholders' conflict and partnership matrix.

Box II.15 - Communication linkage map (example from Lemu Chemeri Peasant Association, Ethiopia)

Box II.16 - Example of a Venn Diagramm for Stakeholder Analysis for an animal health intervention (dipping post, medicine, vaccination, training, water point, vet equipment)

Box II.17 - Stakeholders conflict and partnership matrix (example from Northern Thailand, showing conflicts and partnerships between stakeholders in the use of tree resources)

To prepare the conflict and partnership matrix, group or individual semi-structured interviews can be used to gather the following information:

· Which stakeholder groups have common interests with respect to the planned intervention or technology to be introduced?

· Are there existing partnerships or histories of support and collaboration between or among some of the stakeholder groups? Around which activities, issues or ideals were these partnerships formed? Are there partnerships linked to gender or other group attributes?

· Would it be possible to build upon existing partnerships to implement specific activities? Or could new partnerships be formed?

· Which stakeholder groups have conflicting interests with respect to the interventions in question? Is there a history of conflict between these groups? Are there conflicts linked to gender or other group attributes? How have past conflicts been resolved?

· Are there conflicts so deep and long-standing that certain proposed interventions are doomed to fail?

· Given areas of conflict and partnership, which of the proposed interventions are most likely to succeed?

More information on stakeholder analysis can be found in: Bonnal and Rossi, Guidelines for Participatory Constraints Analysis, op. cit, Wilde, SEAGA Field Handbook, op. cit, Marsden, D., P. Oakley and B. Pratt, Measuring the Process: Guidelines for Evaluating Social Development (1994), INTRAC, Oxford, UK, Carloni, A., and M. Rossi, "Socio-Economic and Production System Study: Background and Materials," in Project Document for Participatory Planning, Field Demonstrations and Sustainable Food Crop Production Intensification to Enhance Food Security in Ethiopia (1995), FAO, Rome, Grimble, R., M.K. Chan. J. Aglionby and J. Quan, Trees and Trade-Offs: A Stakeholder Approach to Natural Resource Management (1995), Gatekeepers Series No. 52, International Institute for Environment and Development (IIED), London, UK, and Guidance Note on How to Do Stakeholder Analysis of Aid Projects and Programmes (1995) Overseas Development Administration, London, U.K.

Quantitative methods

In participatory constraints analysis, quantitative survey methods should be seen first and foremost as a supplement to participatory appraisal methods. If data collected through informal techniques is not statistically precise enough, a questionnaire-type survey may be necessary to estimate crop yields; for example, it may be necessary to complement results from discussions with farmers by a statistical survey, since productivity increases will have to be measured.

The planning stage, beginning with the setting of exact objectives, is one of the most important parts of any survey. In the case of statistical surveys, decisions to be made in the planning stage include the type of data to be collected, the target population, sampling techniques, statistical measures to be used and how to present the results.

Types of Data

The types of data to be considered include:

· continuous data (e.g. weight data)

· counts (e.g. number of individuals in a household)

· scores (e.g. a plant disease score might use numbers of 0 to 5, with zero being healthy and 5 being severely damaged)

· categorical data (e.g. classification of farms into categories such as main crop grown) or binary data (where there are only two categories)

Population and Sampling

Sampling is used where measurement of everything is unnecessary, impossible or too expensive, and just a small fraction of the material is measured instead. In designing a sampling scheme, it is important to define a target population for which information is required. A sample from this population is then selected from which to estimate values for the population (e.g. percentage of farmers using a particular irrigation method).

Two statistical concepts are of importance for sample estimates: bias and precision.

A biased estimate is one that consistently tends to overestimate or understimate the true value for the population. For example, if extension officers are asked to nominate farmers to participate in a survey, they might tend to select the more progressive farmers, or ones who can better afford to take risks. This selection would distort the results. Ideally, samples should allow for unbiased estimates.

Precision relates to the consistency of estimates if the sampling was carried out a number of times. In general, the larger the sample, the more precise the estimate. For example, a sample of 1000 plots will give a more precise estimate of yield than a sample of 10 plots. Precision also depends on the amount of random variation between units. For a given sample size, measurements with large standard deviations will give less precise estimates than those with small standard deviations.

While absence of bias and precision are both important, absence of bias is generally a higher priority. There is little use in getting an answer that is precise but wrong.

The best way to ensure absence of bias is through random selection of units in the population. The overriding principle for selection of a simple random sample is that every unit should have approximately the same chance of being selected. Where this proves impossible, then the target population (and related objectives) may have to be redefined, e.g. farms less than 1 km from a road.

Statistical Measures

There are a number of common statistical measures to be calculated, such as the mean, median, standard deviation and minimum and maximum. In addition, trends and relationships can be studied, and statistical tests carried out. Standard errors are calculated to determine the precision of the estimates.

In the collection of baseline information for demonstration areas, particular attention should be given to avoiding bias when selecting the sample population and to using the most common measures for calculation.

Recording and Presenting Results

In most situations, it will be necessary to design data recording forms (e.g. questionnaires) for each project or demonstration area and technology. Box 11.18 shows an example for recording data on inputs and outputs for a crop demonstration for the SPFS. These forms should be kept to the minimum necessary. Surveys should not be used to collect unnecessary data or information which can better be obtained using participatory methods. In using data recording forms, copying by hand should be limited as much as possible in order to avoid the possible introduction of error and loss of time.

Today, the use of computers and statistical programmes is fairly standard for the analysis of quantitative data. In addition to common data base and statistical computer programmes such as Dbase, SPSS, and SAS, spreadsheet programmes can also be useful, in particular if data bases are relatively small and the statistical calculations to be carried out are not very complex.

Although often overlooked, the presentation of results is an important part of every statistical analysis. Well-presented results can greatly facilitate the interpretation of statistical data and the drawing of conclusions. Numerical results are usually presented in the form of tables or graphs, depending on whether numerical precision is required or only an indication of trends. Some of the statistical data may feed directly into the financial calculation of farm budgets.

Steps to Follow

The preparation and conduct of quantitative surveys generally involve the following steps:

· decide on objectives of survey
· identify exactly what data are needed
· decide on target population
· determine techniques for obtaining an unbiased sample of the target population
· decide on what statistical measures will be used, and how the results should be presented
· prepare data recording forms/questionnaires
· test the questionnaires
· identify and train enumerators
· carry out the data collection exercises (supervision is important if enumerators will collect the data)
· after completion of field work, check the questionnaires and code responses
· transfer the data to processing medium (i.e., computer)
· check for errors and validate data
· carry out the actual analysis (i.e., calculation of measures)
· interpret and present the results

More information about quantitative survey methods can be found in: Sherington, J., Statistical Concepts in Research (unpublished paper prepared for NRI's Grain Storage Management Course) (1997), Natural Resources Institute (NRI), Chatham, UK, and Farm Demonstration Protocol Guide, (1997), SPFS/DOC/20, FAO, Rome.

Box II.18 Format for recording data on inputs and outputs for a crop demonstration

Name of farmer: _____________________________________
Location of demonstration: ____________________________
Season (e.g. 1997/98): _________________________________

Tick the appropriate box to indicate whether a field or plot within a field was used:

Field: Plot:





Labour (work days)

Unit Cost

Hired Labour cost

Labour (work days)

Unit Cost

Hired Labour cost
























Land preparation

Nursery planting




Fertilizer and manure application

Water application

Pesticide application






Other (please specify)



Plot Crop






Unit price




Unit price


Crop 1








Crop 2








Crop 3

Nursery plants







The above table is an example of the data sheet required for a form where the demonstration is conducted on more than one crop (e.g. with intercropping or crop rotations).






Area grown



Unit price


Area grown



Unit price


Crop 1

Crop 1 residues*

Crop 2

Crop 2 residues

Crop 3

Crop 3 residues


* Residues refer to straws which may be sold for plant nutrition, housing, animal feed or fuel











A. Labour Expenses

A1. Male hired

A2. Female hired

A.3 Family labour (workdays)

B. Non-Labour Expenses

B1. Crop 1

B2. Crop 2

B3 Crop 3

C. Total Variable Cost (A+B)

D. Returns to Production

D1. Produce sold

D2. Produce used on farm for home consumption

Gross Margin (D-C)

Gross Margin to Labour (D-C)/A.3