Chapter 1

To RAP or Not to RAP (and the Basic Concepts)

Main Points

1. RAP is intensive, team-based ethnographic inquiry using triangulation, iterative data analysis, and additional data collection to quickly develop a preliminary understanding of a situation from the insider's perspective.
2. The phrase "Rapid Assessment Process" defines the methodology and the acronym, RAP, communicates the essential ingredient for successful implementation.
3. RAP allows a team of at least two individuals to quickly gain sufficient understanding of a situation to make preliminary decisions for the design and implementation of applied activities or additional research.
4. Results can be produced in as few as four days, but implementation more typically requires several weeks.
5. RAP uses the techniques and shares many of the characteristics of ethnography, but differs in two important ways: (1) more than one researcher is always involved in data collection and teamwork is essential for data triangulation; (2) more than one researcher is involved in an iterative approach to data analysis and additional data collection.
6. The intensive teamwork for data collection and analysis is an alternative to prolonged fieldwork and produces results that provide insights into the perspective and worldly view of the participants in the local system.
7. RAP is especially appropriate for a variety of situations where qualitative research is needed.
8. RAP can be used for monitoring and evaluation.
9. Sometimes survey research is not an option for initial research because not enough is known to prepare the questionnaire.

Example One: Student Services at a Community College

The new dean at a community college in the Pacific Northwest was made to understand during her interview for the position that there was serious discord in the Student Services Division. After being offered the position, Pat (a pseudonym) was told that she was expected to contribute to a "healing" process. By November of Pat’s first term, she realized the rift was greater than she had anticipated, but that many of her actions to bring about change were being warmly received by many in the division. Pat knew that there were no easy solutions to the organizational culture issues in the division, and hoped that an examination of this culture could contribute to reconciliation and help individuals refocus on the mission of the division and the college. Pat instinctively knew that the history of the different conflicts was too long and the issues too complex for research based on a set of questions to be administered to everyone. She assumed that there would be a very low level of participation in research based on questionnaires, even if appropriate questions could be identified. Pat also knew that the community college did not have the luxury of the extensive time and other resources required for traditional qualitative research. At this point, she requested a Rapid Assessment Process of the organizational culture of the division. She knew that the goal of RAP was to listen to the stories of individuals involved in an organization and from these stories to quickly identify common themes.

Most of the participants’ comments were grouped into categories and identified as constraints to the ability of the people involved in the Student Services Division to do the best job possible. There was a general consensus that performance was not always as good as it should be, and that most people truly wanted to do better, but that there were a range of interrelated factors that prevented this from happening. The six constraints identified involved: (1) communication; (2) physical space; (3) technology; (4) utilization of people’s time, talents, and creativity; (5) increases in the number and complexity of regulations; and (6) inadequate resources.

While there was general agreement that the purpose of working in student services was to provide services to students, there was significant disagreement about the characteristics of the student population and what it meant to serve them. "Students are the reason why I'm here" was a view expressed by almost everyone interviewed. Several individuals, however, objected to the use of the word "service" to describe the relationship with students. One participant suggested that students view services as things they are entitled to and people who provide services as their servants. For some participants, the defining characteristic of service was being "available." This was identified as being an especially important aspect of service for "walk-in traffic." Most participants appeared to recognize a difference between students needing merely "regular" services, such as academic advising, and "high-maintenance" students needing help with issues such as immigration/visa problems and emotional and mental-health problems.

Several participants consistently used the words "customers" or "clients" instead of "students." There was significant disagreement about the characteristics of the student population and how this had changed over time. Some indicated that "our students have always had special needs," while others pointed to a growing trend in "emergent situational disorders that happen as a result of overstress in personal academic life choices." A few participants noted their students were less prepared for college now than students in the past had been; one participant said, "students do not comprehend written communication and have trouble with verbal communication." Another participant noted, "We are doing more handholding than we used to do." One person, who had worked at the college for over three decades, said there had been a change in general life experiences that affected preparedness, especially with regard to readiness in a professional/technical arena. Another participant shared a story told to him by one of the technical instructors of a student who was asked to get a Phillips screwdriver from the tool crib, and who returned empty-handed, saying that all he could find was a "Stanley."

Several participants expressed surprise that changes in the characteristics of the student population were even being discussed. "I wasn’t aware that there are major changes" in the student population, one participant commented. One participant suggested that perhaps the changes were not in the students, but in the staff, saying that the staff had grown older and that maturity had changed perceptions. "I’m not the bleeding heart liberal I was when I started," this participant added. Some students were described as aggressive and "bullish" in demanding services from staff, "expecting everything to be given to them and believing it’s always someone else’s fault when things went wrong." Other students were described as lacking sufficient skills to request services they needed and were entitled to. The argument was that, for some students, services should include advocacy on behalf of the students. An individual dealing with special population students offered a plea for greater effort be made to ensure that "no one drops between the cracks" and that his colleagues should realize that this request was not for special advantages, but for a "level playing field."

Teams were organized around the constraints identified by the RAP and everyone working in the division was asked to participate on at least one team. Funding was identified that teams could apply for to address some of the constraints. The identification of serious differences in how individuals who work in the division understand the terms "students" and "services" has made possible constructive dialogue between the different factions.

Example Two: Village in the Western Sudan

The soils are sandy, the rain sparse and unpredictable, and the temperatures often above 110^oF. at the edge of the Sahara Desert in the western Sudan. Here, rural families survive during the better years through a combination of agriculture, livestock, and submission to the will of Allah. Margins for error are extremely slim. The adoption of an inappropriate agricultural intervention could mean ruin for the family and irreversible damage to the environment.

In the early 1980s the U.S. Agency for International Development was providing assistance to Sudan to establish an agricultural research center in western Sudan to help address these issues. The survival strategies of rural households in that environment were not understood. Mohamed el Obeid, a Sudanese agricultural development specialist and I spent one week in 1982 in a village northwest of al Ubayyid (el Obeid), the provisional capital of North Kordofan (Beebe 1982). My objective was to experiment with the new research methodology called "Rapid Appraisal." The two weeks in the village were amazing. We spent our days talking with groups of farmers or individuals. We conducted interviews about farming practices in the farmers' fields. We had extensive conversations with the village religious leader and the owner of the only shop (one of only two structures made of sun-dried bricks in a village where all the other structures were made of grain stalks and straw). We visited a slightly larger village where farmers could sell grain and other agricultural products. We spent our nights trying to figure out what we had learned and were often joined by men from the village who used our presence as an opportunity to discuss life in general, including the plans of one of the slightly more prosperous inhabitants to take a second wife. The structure imposed by the Rapid Appraisal methodology allowed us to quickly understand some of the important concerns of the residents of the area. Prior to our visit to this village, the assumption of both Sudanese and American agricultural research scientists had been that individual farmers were free to move at any time between gum arabic production (using the nitrogen-fixing tree Acacia senegal) and field crops such as sorghum and millet. Attention to their descriptions of crop rotations, especially when this was discussed with farmers in their fields, where crops could be observed, suggested that decisions by a farmer’s neighbors could be a significant constraint. If a neighbor's large gum arabic trees, which would harbor birds, were too near, this would prevent the farmer from planting field crops until the neighbor was also ready to cut down his gum arabic trees and plant field crops. Subsequent research by Reeves and Frankenburger (1981) on agricultural practices in North Kordofan showed that adjacent fields were often cultivated by farmers who were related and that this also played a role in timing the rotation of a field from gum arabic production to field crops. Recognition of constraints on the decision-making abilities of individual farmers had a significant impact on the approach to farming systems research and extension proposed for the western Sudan.

The Need for the Insider’s Perspective

Despite the many kilometers that separate the village in western Sudan and the community college in eastern Washington, the two situations share an important characteristic. Both are complicated situations where initially not enough was known to develop a questionnaire. Only the insiders in each of these situations were in a position to define the elements of their systems and identify those elements that were most relevant to the issues they faced. The insiders in the Sudanese village knew at least intuitively that they could not change their crop-rotation pattern without regard to the actions of their neighbors, but it is unlikely that a question could have been formulated in advance that would have elicited this information. Likewise, it is unlikely that a question could have been developed to elicit information on the different ways staff at the community college defined students and services, since there was no reason for outsiders (or even many of the insiders) to think this was an issue. Another characteristic shared by these two situations was that results were needed quickly and that, even if the results had not been needed quickly, there were not sufficient resources for traditional, long-term fieldwork. The approach to research that focuses on getting the insider’s perspective is referred to as "qualitative" research or inquiry. It is useful under some circumstances to differentiate between qualitative research and an approach to qualitative research called "ethnographic" research. They are, however, often used interchangeably. In chapter 2, I will return to the relationship between the insider’s perspective and ethnographic/qualitative research (see ethnography, p. XXX; Emic and Etic, p. XXX).

Rapid Assessment Process and Intensive Team Interaction

RAPs similar to what was done at the community college in the Pacific Northwest and in the village in western Sudan provide a way to investigate complicated situations in which issues are not yet well defined and where there is not sufficient time (or other resources) for long-term ethnographic research. RAP shares many of the characteristics of ethnographic research. RAP, however, substitutes intensive, team interaction in both the collection and analysis of data for the prolonged fieldwork normally associated with ethnography. RAP will produce solid qualitative results that can be expected to be different from those produced by longer-term fieldwork. In some cases, intensive team interaction over a short period may produce better results than a lone researcher over a long period. RAP will almost always produce results in a fraction of the time and at less cost than traditional ethnography.

Basic Concepts

RAP uses the techniques and shares many of the characteristics of ethnography, but differs in two important ways: (1) more than one researcher is always involved in data collection and the teamwork is essential for data triangulation; (2) more than one researcher is involved in an iterative approach to data analysis and additional data collection. The intensive teamwork for both the data collection and analysis is an alternative to prolonged fieldwork and produces qualitative results. RAP allows a team of at least two individuals to quickly gain sufficient understanding of a situation to make preliminary decisions for the design and implementation of applied activities or additional research. Results can be produced in one to six weeks. While the time period for RAP is recognized as arbitrary, my experience has convinced me that a minimum of four days is required for iterative data analysis and additional data collection, an issue I will return to in subsequent chapters (see Too Little or Too Much Time, p. XXX).

The two basic concepts of RAP define its relationship with traditional ethnography and allow results to be produced quickly. The two basic concepts are

1. intensive teamwork as part of the triangulation of data collection, and
2. intensive teamwork during the iterative process of data analysis and additional data collection.

Adherence to these concepts can provide a flexible but rigorous approach to the rapid collection and analysis of data.

Relationship of the Basic Concepts to Research Techniques

RAP is defined by its two basic concepts instead of by a specific set of research techniques. While, traditionally, some research techniques have been associated with rapid research methods, these methods are not necessarily required. Specific research techniques for use in a given RAP are chosen from among a wide range of techniques available to qualitative researchers and are chosen based on the specific topic being investigated and the resources available to the team. Specific techniques used in a RAP can vary significantly depending on the situation.

<fig1.1>

Figure 1.1 illustrates the relationship of the basic concepts and illustrative research techniques associated with them. As noted above, the listed research techniques are not the only way of achieving the basic concepts, but are techniques that have been found to work together under some field conditions.

Conditions Where RAP Is Especially Appropriate

I have identified, as the types of situation where RAP may be the most appropriate methodology, complex situations where the categories and words used by the local people involved in a situation are not known.. I will try to make the case throughout this book that if results are needed immediately, RAP may be the only choice. RAP is also appropriate for a variety of situations where qualitative research is needed, even if there is no time constraint. Creswell (1998, 17–18) has identified what he calls compelling reasons for undertaking a qualitative study as opposed to a quantitative study. These reasons can help identify situations where RAP is an appropriate approach for research:

1. When a topic needs to be explored. This occurs when there is insufficient information available to identify variables or to know how local people identify them.
2. When the question begins with "how" or "what." These are the questions when the critical elements in a situation and their relationship to each other cannot be identified.
3. When there is a need for a detailed view. Situations may be so site specific that general information, especially information that covers a larger area or very many people, may not provide insight into the issue being investigated. There are times when only detailed information on a limited topic is useful.
4. When there is a need to study individuals in their natural setting. Without access to the environment of the person providing answers, it may not be possible to know what questions should be asked or to understand the context for the answers given. If participants are removed from their setting, it may lead to contrived findings that are out of context.
5. When there is a need to emphasize the researcher's role as a partner instead of as an expert. There is growing recognition that research needs to be collaborative with all parties that stand to benefit. People have limited tolerance for experts who pass judgment on them. People increasingly are demanding that they be allowed to articulate their own stories.

RAP is not an appropriate methodology if quantifiable results are needed, such as the percent of individuals, opinions, and homes in different categories. RAP may be the appropriate methodology for identifying the most relevant categories, and the most appropriate labels for these categories, but the research to collect and analyze such data is usually not a part of RAP.

Using RAP for Monitoring, Evaluation, and Midcourse Corrections

RAP can be used for monitoring and evaluation. The identification of specific midcourse corrections during the implementation of an activity is another task for which a RAP may be useful. When ongoing monitoring of an activity suggests problems with implementation and the causes of these problems are not obvious, a RAP team can explore questions as fundamental as whether the local people and the parties responsible for the activity agree on what constitutes success and failure. A RAP approach is especially useful in identifying the unexpected. A report based on a few weeks work and delivered immediately allows for midcourse corrections. A report prepared by a team in a situation where local people have been full partners increases the chances that recommendations for changes can be implemented, and increases the opportunities to implement changes, even before the recommendations have been made formally.

When faced with the limitations of time and resources, the temptation can be to make a very quick visit with the most easily reached local participants. This is sometimes referred to as "research tourism." Another temptation is to do a questionnaire survey even when there is agreement that the most important issues have not yet been identified, and that the categories and words with the greatest relevance to the local people are not known. The rationale seems to be that something needs to be done, and that anything is better than either research tourism or nothing.

When Survey Research Is Not the Best Option

The story of the two neighboring villages in Africa, one where almost all the babies under one year of age were boys and another where almost all were girls, has become part of the folklore of health-care development workers. The villages were identified as a result of a survey of health and mothers’ knowledge of health-care practices for children under one. The storyteller usually relates, empathetically, that the study was funded by a major international donor and implemented by a professional researcher who had carefully prepared the questionnaire and trained and supervised the field staff who carried out the interviews. The results were so surprising that local ministry personnel were sent to the villages to investigate. What they reported after their visits to these villages is more interesting than the original "results." The mothers in the first village considered boy babies much more desirable than girl babies and, when asked about their children, tended not to mention the girls. Thus the village appeared to have almost no girls. The mothers in the second village also considered boy babies more desirable than girl babies. However, in this village mothers did not tell strangers about their boys out of fear that if they brought attention to their sons, harm would seek out the boys. Thus the village appeared to have almost no boys. Like so much folklore, it is not possible to identify the source of this story.

Survey research based on questionnaires, a group of written questions to which individuals respond, have been used and misused worldwide. When faced with the need for information about situations, researchers have often tended to do a survey. Survey research has been viewed as reliable, producing similar answers every time the questionnaire is administered, and relatively quick when compared to traditional ethnographic research. My argument is that often survey research is not an option for initial research because not enough is known to prepare the questionnaire. To prepare a questionnaire, you need to be able to identify the relevant elements of a situation, the specific categories that are important to the respondents, and the words they use for these categories. Since a questionnaire cannot identify unanticipated, site-specific relationships, it is limited to validating relationships articulated in advance.

Unless questionnaire survey research is based on the categories and vocabulary of the respondents and the context of the data is understood, the results may not be valid measures of what they purport to measure. Such results can be reliable without being valid, since different researchers administering similar questionnaires would likely get the same results.

An experiment by Stone and Campbell designed to examine the accuracy of practices, attitudes, and knowledge (PAK) surveys concerning fertility and family planning in Nepal illustrates this. Stone and Campbell hired and trained interviewers to administer the Nepal Fertility Survey to women in three villages. They then cross-checked the information on the survey forms by using other methods, including casual conversations and unstructured interviews. During the survey, 36 percent of the respondents claimed they had not heard of abortion. When Stone and Campbell asked about awareness of abortion, 100 percent knew about abortions and even "maintained that is was inconceivable that someone had not heard of [it]." Stone and Campbell suggest that part of the explanation is that abortion is considered a "religious sin" and that some respondents were "insulted by the question" (Stone and Campbell 1984, 31). When they talked to the respondents who had indicated during the survey that they had not heard of abortion, every respondent had reinterpreted the question to make it more threatening. They found that respondents had interpreted the question on whether they had "heard of abortion" as a question on knowledge of technique or knowledge of who had had an abortion. They found that every woman who had reported little knowledge of family planning in the survey reported that they had difficulty understanding the questions and had been embarrassed by them.

Stone and Campbell identify cultural reinterpretation and problems of context as the factors that influenced the results and note that these can be problems for research done in the United States as well as research done overseas.

Specific problems have been identified with survey research on sexual behavior, voting, and geographical knowledge that may be relevant to other survey research as well. Clement suggests that there are two sources of error, invalid answers and volunteer bias, for research about sexual behavior. He notes that the validity of answers depends upon the ability of respondents to remember and their readiness to share the information, and that these are influenced by how the questions are posed (Clement 1990, 46). He specifically notes problems with crosscultural research, including research within the United States but at different universities or with different ethnic groups. Despite increases in the sophistication of survey research methodology and analysis for political/voting surveys, results have become less accurate. More and more individuals are simply refusing to provide answers to pollsters and increasing numbers identify themselves as "undecided," even when they have decided. Surveys on respondents’ knowledge of geography have found "astonishing geographic ignorance." Phillips (1993) argues that these finding are the result of asking the wrong question. He notes that the questions are usually based on the categories used by geographers and not the mental images of people whose images are not "map-like." Phillips argues for the increased sensitivity to the nature of geographic mental representation as a basis to evaluate geographic knowledge. In all three of these examples, questions that fail to consider the cultural context with specific attention to the definitions and categories of the respondents have produced answers with limited validity.

It is sometimes incorrectly argued that survey research is quicker and can be done with less-experienced, less-qualified researchers, compared with RAP. Data collection by survey sometimes requires less time, but data analysis almost always takes more time. Data usually must be coded, entered into a computer, and then analyzed in separate steps and at places removed from the research site. Survey enumerators may have to make fewer independent decisions than a qualitative researcher does, but good survey research cannot be carried out without training and close field supervision. In addition, special training in instrument design and data management ensures that survey research usually does not include local participants as full members on the research team (Chambers 1991, 526).

RAP may identify the need for questionnaire survey research to supplement its results. As noted above, RAP cannot provide information on the percentages of respondents in different categories, and this information can be critical. Survey research can provide this type of information. However, RAP can provide the categories, vocabulary, and context necessary for the preparation of the questionnaire. The argument here is not against using questionnaire surveys, but against using them as the first step for trying to understand complex situations before local categories are known. In addition, a RAP may be a better starting point for some research because of its ability to discover relationships within the situation that may not have been anticipated. The use of techniques associated with RAP does not guarantee success in identifying important relationships, but initial research on complex situations based on a questionnaire often ensures that they will be missed.

The Need for Caution about the Use of RAP

Robert Chambers's observation concerning Rapid Appraisal also applies to RAP, in that there is a danger it "could be over-sold, too rapidly adopted, badly done, and then discredited, to suffer an undeserved, premature burial as has occurred with other innovative research approaches" (Chamber 1991, 531). When numerical data is needed, RAP by itself will probably be an inappropriate methodology, but it might contribute to the design of a survey questionnaire to collect numerical data. When situations are especially complex or when an entire cycle, such as a growing season or a school year, needs to be investigated, long-term qualitative research may be necessary. There may be situations where it is culturally inappropriate for a team of researchers to interview an individual. Other valid reasons for concern about RAP include spending too little time on the activity, failure to consider the political and economic context, problems with team composition, choice of respondents and informants, and a failure to recognize a difference in power between the team and the local community. To date there has been a general lack of confirmation of RAP findings. RAP, along with other qualitative research methods, lacks credibility with some funding agencies, while others funding agencies have very unrealistic expectations about what RAP can accomplish and sometimes pressure researchers to do RAPs in inappropriate situations. These issues will be discussed in more detail in chapter 5 (see Problems with Credibility, p. XXX).

In the next chapter I will discuss the first of the two basic concepts, the one relating to data collection, triangulation, and intensive teamwork. Because of the importance of ethnography/qualitative research to RAP, I will spend time exploring it. However, since this is not intended to be a book about ethnography and qualitative research, issues can only be introduced and you are encouraged to refer to the additional readings listed at the end of the chapter for more information. The specific techniques that are introduced are those that have proven to be most relevant to RAP. Most of the techniques are designed to help facilitate the telling of stories as opposed to the eliciting of answers. Others are designed to ensure that the RAP team records data in ways that will make the data useful and easier to analyze. A specific RAP may use only a few of these techniques and may use other techniques that are not covered. One of the strengths of RAP is that it is not based on the use of a specific list of techniques.

Additional Readings

The readings listed below are some of the most-often cited references dealing with rapid research methods. Even though Scrimshaw and Gleason (1992) focuses on health programs, material in this book will be useful to researchers from a variety of fields. The full text of the Scrimshaw and Gleason book (as well as several other publications on rapid research methods) is available on-line at <http://www.unu.edu/unupress/food/foodnutrition,html>.

Van Willigen, J., and T. L. Finan. 1991. Soundings: rapid and reliable research methods for practicing anthropologists. Washington, D.C.: American Anthropological Association.