II. DEFINITION OF THE RRA TOOLKIT

2.1 Of What Does the Toolkit Consist?

Since this is a review paper rather than an original work, massive liberties have been taken with the available materials on RRA. For this section, I have adapted a table (Figure 1) from an extensive treatment of RRA methods (Grandstaff and Grandstaff, page 84, in Khon Kaen University, 1987), which itemizes the gamut of tools and their uses. The toolkit, in other words, is both eclectic and broad-ranging. This table summarizes well, however, what types of techniques fall under the heading of RRA methods in a wide number of approaches.

The tools include:

1) interview and question-design techniques for individual, household, and key informant interviews,

2) methods of cross-checking information from different sources,

3) sampling techniques that can be adapted to a particular objective_,

4) methods of obtaining quantitative data in a short time frame,

5) group interview techniques, including focus-group interviewing,

6) methods of direct observation at site level, and

7) use of secondary data sources.

RRA approaches combine selected elements of the toolkit to provide a comprehensive methodological framework for gathering and assessing information in the field. The typical length of time that an RRA exercise is allocated ranges from 3 days to 2 months and can cover single villages or a 10,000 square mile project area. The methodological choices enter in when deciding who should be on a team, whether to collect quantitative or qualitative information about a particular issue, on whether to apply formal survey methods to field interviews, when selecting the range of individuals and groups to talk to, when selecting the representative sites to be seen, when allocating tasks among team members, and when making use of interactive tools.

There is no set way to combine these tools for the purposes of a study or planning exercise. In some circumstances, the choice of tools is determined by an already defined set of parameters. Let us assume, for instance, that information on the local land-use system, local needs, and institutional capacity is needed quickly to help in preparing a preliminary design for a social forestry program for a sizable province. The concerned planner has allocated enough resources for three social forestry professionals from varied disciplines and three local extension agents (foresters and agriculturalists) to spend two weeks in the province to gather some additional information to what is available in secondary sources and what is presently known to staff of the two extension services. In this situation, the team is a given, the time frame is a given, and what remains to be determined is how many and which people will be interviewed, how many and which sites in the province will be visited, what forms of measurement and direct observation will be employed, and whether any interactive data gathering techniques will be used, such as mapping or ranking of priorities with local villagers and/or extension staff.

In other circumstances, decision-making may be relatively open, within the confines of a six-months to a one-year deadline, or within a budgetary constraint of only 1% of the budget for feasibility studies. For example, a watershed program is being developed for some pilot watersheds in a province and planners need to understand the present land-use system from the multiple perspectives of the provincial and local government and the local population. Information is needed within a six-month period and any team can be constituted within reason. Here, the choice of expertise and number of team members is fluid and the research can be done over a one-week, two-week, or two-month field visit.

There are a number of what are termed here as "package approaches" that have been developed by various RRA experts which provide the potential user with a relatively set combination of tools, often geared to a particular purpose --in-depth problem analysis, general program formulation, or village-level planning of activities. These include, for example:

1. Agro-ecosystems analysis at IIED

2. Farming systems research diagnostic surveys at the CGIARs

3. Diagnosis and design at ICRAF

4. Village dialogue, Autopromotion, Microplanning, and other village-level planning approaches

5. RRA as promoted at IDS, Sussex

6. RRA as promoted at Khon Kaen University, Thailand

One problem for institutions trying to develop a host-country expertise in RRA methods is that the similarities and differences among these approaches are not clearly understood by newcomers to this field. In some countries, individuals from the same government agency have received training at different points in time from trainers advocating different approaches to what is basically RRA (sometimes called RRA and sometimes, diagnostic survey approaches). Yet, because the agency staff never really understood that these were different aspects of the same type of methodology, the trainees of the two different courses believed they were doing two different things, impeding the trading of information among the two sets of participants. Those trainees who had participated in both training courses were unaware the courses were closely related and compartmentalized the information from each course, rather than having each set of coursework reinforce the other. What is different in these approaches is often the choices of resources, time frame, or mix of interview types, not the basic principles or underlying methodological rules. This should become clearer in the section of the paper describing the elements of the toolkit.

Figure 1: TOOLKIT FOR RRA*

General Methods	Semi-Structured Interviews			Group-Level Interviews and Observation
Interview Guides	(HH or key informant)			(Village or Site level)
	Team Dynamics Group/Verbal	Individual Tools Information	Logical Schematics Time/Space Direct Observations	Group/Verbal	Pre Existing Information	Direct Observations
Triangulation (crosscheck)	Sondeo (rotating pairs)	- Non-leading questions	- Crop calendar	- Focus group	- Studies	- Roads/trails
Secondary Data	Protocols	- Six helpers who, what, when where, how, why	- Timeline	- Village meeting	- Reports	- Village walks
Team Interaction	Minimum Data Sets	- Probing	- Cross section Village/Resource Transect	- Open discussion	- Site data;	- Markets Shops
		- Local Names		-Discuss survey results	Census;	- High place w/view
		- Folk Taxonomy (ITK)		- EIA in village	Statistics;	- Houses, compounds
		- Oral History	Historical T.	- Staff interviews	Maps;	- Meeting places
		- Case History			Aerial Photos Spot Imagery.	- Festivals
						- Stay overnight
					- Expert opinion
Direct Measurement
Photos						Counts
Tape Measure						- Houses
Tape Measure						- Trees
Scales						- Fields
Volume Measure						- Vehicles
Pedometer					Indicators	- Bicycles
Altimeter					- General	- Tractors
Levels					- Specific	- Ponds
					- Locally derived
					- Dynamic over time

* Adapted by author from Grandstaff and Grandstaff (referenced in text), representing variety of tools used in RRA at Khon Kaen University, Thailand

2.2 A Tentative Framework for the Application of the Toolkit

Figure 2 provides a tentative outline for the kinds of purposes for which RRA methods can be used and outlines the types of methods in that toolkit which are most appropriate for each of these different purposes. This is not a cut-and-dried set of choices -- but depends upon resources, objectives, and the time period within which the RRA exercise will be carried out.

One confusion that arises in discussions of the toolkit is between the use of a tool as a means of information gathering, and analytical assessment versus its use as a means of presenting a team's findings. For instance, the farming systems research transects of land-use are used both as an interactive tool for letting interviewees understand what researchers wish to know about their land-use system, also, for analyzing the nature of issues and problems entailed in that system, as well as to present information in a concise and pointed fashion to decision-makers about the findings. These multiple purposes become confused in methodological discussions of whether or not and how a tool should be applied. The reader should keep this point clearly in mind in the section that compares different approaches, since the same tools can serve different purposes even within the same RRA approach. While a tool can be a powerful means of presenting information, it may or may not be the most sound tool for gathering or evaluating that piece of information in terms of the ,analytical problem or topic that the RRA sets out to assess. Those new to RRA often make the mistake of assuming their information is correct if it is collected in a systematic table as outlined in a particular field manual. This s not necessarily the case. For example, simply making use of a village transect to collect information does not guarantee that the team will correctly identify the full range of cropping strategies in each altitudinal category. Nor does use of a tool as an analytical aid necessarily mean it will be the best y of introducing findings into a planning session.

Figure 2: Possible Use of the toolkit

Fig. 2.1a

	Purpose	Required Time Frame	Team Composition	Cross-checking	Group vs House Interviews
1.	Initial program formulation: - define problem parameters, possible interventions, key issues, and information needs.	1-2 months infield and/or including literature review.	Small multi-disciplinary groups of experts, local extension agents, officials, and villagers.	Direct observation, key informants, group interviews, secondary data & topographical maps. Team subgroups rotate members to maximize perpectives.	Maily group interviews team and groups. HHS visits and interviews help identify special interest groups, key informants. Creative mix of these decided during field exercise and changed in light of emerging information.
	IdentificationStage		Subdivide as needed a la Conway, Hildebrand, or ICRAF.
2.	Indepth look at specific problem areas at design stage	1/2-10 days per representative area, total: 1 week-2 months.	Smaller team of experts, agency staff, facilitators, villagers, NGOs.	Reasking for information in different ways during interviews. Situational analysis, Hypothetical questions, Participant observation, Legal statutes/ casebooks.	More use of focus groups, household interviews, probing, six helpers, key informants. May attempt more systematic combination of interviews than in (1) above and quantify to greater extent. More focus on groups. Household interviews to explore individual strategies--perhaps with well-defined sample. Focus groups useful.
3.	Appraising esign to help finalize project plan	1-2 months	Mix of subteams focussing on different sets of key issues. Many team meetings to exchange information.	Ask simple questions in each site. Combine direct observation, interviewing and Key informants as in 1.	More focus on groups. Household interviews to explore individual strategies--perhaps with well-defined sample. Focus groups useful.
4.	Initial stages of implementation: - Help define monitoring system - Explore key issues	1 month	Multi-stage exercise		Interview villagers to see what they would like to know. Interview groups of field staff of different ranks to determine how much data they can collect.
	- Village level planning process Interactive Planning	1/2 day-1 week per village. May be in stages rather than continuous period.	Project staff, trainers, facilitators, key experts, NGOs, villagers.	Compare answers from different sessions. Do HH interviews to elicit a wider range of opinions, use games to elicit local thinking/knowledge.	Multistage interviews with whole group and special interest groups allow villagers to think about problems between sessions and allow "invisible" groups to speak up.

Fig. 2.1b

	Sampling Techniques	Data Over Time	Interactive Toolkit	Minimum Data Sets	Possible Follow-on
1.	Introduce random site selection. Random choice of informants halfway through survey to check hypotheses. Null hypothesis	Change in land use, prices, markets, cropping, resources, employment, migration. Cycles: - climate - rain - life cycle of house hold	Limited use due to time constraints	- Open-ended checklists.	- Formal random surveys of population, family system, land use. RRA of special issue. Indepth village study of representative site conditions.
				- Agroecosystems checklists.
				- Diagnostic and design check list.
				- Khon Kaen checklist.
2.				Agroecosystem Design.	Economical design of a formal survey of key issues perhaps using results of RRA to define sample size based on variation in local population.
				Khon Kaen checklist.
				Physical measurements devised by RRA practitioners.
3.	Null hypothesis Stratified sampling by site and whole project area	Already available to team from earlier RRA and formal study.	Like 1 --unlikely to be an important tool.	Coordinate checklists among subteams.	ME and other studies during implementation.
				Checklists on institutional capacity are key here.
4.				Evaluate possible checklists by pretesting in household and group interviews in field.	Devise ways for communities to monitor progress as well as monitoring by agency itself. Formal surveys.
		Collect same as above if not already available	All tools to help village analyze existing contraints & potential victim. Sketch maps for village Land-use plan. Labor calendar to identify labor shortages.

Fig. 2.2a

	Purpose	Required Time Frame	Team Composition	Cross-checking	Group vs House Interviews
5.	Monitoring and valuation during implementation	Follow general methods outlined for in-depth study at design stage
	- Problem analysis for qualitative issues
	- Cross check formal survey data	1-3 weeks	Small team of technical staff, research technicians,social scientist, field staff, M&E unit members, and villagers.	See lists above in 1. Can make extensive use of RRA physical measurement techniques.	Discuss reasons for findings from formal surveys with target group/clients.
	- Key planners into areas of implementation for which special study is needed.	1 day - 1 week	M&E staff, field staff, selected planner(s)		interview special interest groups.
6.	Mid way assessment of project focus and achievements	1-2 months in field and/or including secondary source review	Experts, project staff, and officials. Subdivide to draw off officials in discussions where their presence would bias information.	Spot visits apart from planned schedule. Cross check group interview results with HH and focus group interviews. Observe achievements in forest and farmer lands. Discuss process there not elsewhere.	Both group and household interviews, particularly including special interest groups. Focus group and informal discussion with field staff of various ranks so underlings are free to express opinions.
7.	Final evaluation	1-2 months	Same as above	Cross check M&E results, special study data.	Same as mid way review. Be sure to do some probing and use helpers in interviews where time permits.
8.	Other special uses Tools to present data in convincing form to planners		Have planners join teams for a few field visits to participate in the appraisal and interview sessions.		Problem-solving sessions with villagers in group interviews.

Fig. 2.2b

	Sampling Techniques	Data Over Time	Interactive Toolkit	Minimum Data Sets	Possible Follow-on
5.	Random area selection by stratified choice of sites. Null hypothesis	Same lists as above in 1	Can make use of this particularly to analyze data from local perspective and to convince managers.	Can follow checklists fairly closely if based on good understanding gained in design stages and pre-project surveys.	Beneficiary assessment as outlined in World Bank document by Larry Salmen.
	Random crosscheck of sample findings. Concentrate sample on special interest groups for which data is least available	Same as above		Thus elicit comparable data for a number of different sites.
	Same as above	Same as above
6.	Random site selection in addition to suggested sites. Draw sample from ME formal survey sample for interviews. Use health or political or nursery lists to pick random sample	Same as above	Labor calendar to identify and time constraints Ranking games to identify fits of technical packages to local needs and interests	Select criteria such identified in Parker evaluation to evaluate objectives. Select items on checklist based on hypothesis to test and problems expected. Clearly allocate responsibility to team members where they need to interact among technical disciplines or collect information for each other.	Small sample survey
7.	Subcontract small formal survey of key topics during eval. period. Null hypothesis. Same as above.	Same as above	Village transect may help evaluate on which Lands farmers adopting different interventions
8.	Provide quantitative evidence of key "surprising" findings through a form of sample surveying to complement the RRA itself.	Same as above	Overlays of color-coded information of different factors on sketch maps, village transects, time calendars, flow	Draw planners into discussions to devide on checklists and teams' methodology.