Key Words: Q-Methodology, participant perspectives, national forest management

Toddi A. Steelman is Assistant Professor of Environmental and Natural Resource Policy at the University of Colorado at Denver’s Graduate School of Public Affairs.

Lynn A Maguire is Associate Professor of the Practice of Environmental Management and Director of Professional Studies at Duke University’s Nicholas School of the Environment.

*Both authors would like to thank Steve Brown, Ron Brunner, Dan Durning, Daniel Fiorino, John Hird, Frank Laird and Andrew Willard for their extensive and helpful comments on an earlier draft of this paper. We would also like to thank Mary Matella and Karin Maloney for research assistance.

Since the National Forest Management Act (NFMA) was passed in 1976, National Forest planning has left many expectations unfulfilled. NFMA sparked a revolution in forest planning and has been cited as being the "most detailed and participatory forest and rangeland process ever undertaken" (Gippert and DeWitte, 1996). While the desire to incorporate the public’s concerns into the planning process was well intended, crafting management plans that satisfy the many constituencies served by National Forests has proven a contentious task. Since NFMA became effective in 1983, there have been over 1,200 administrative appeals of the 124 forest plans (USDA, 1997). The consequences of failing to produce policy that is acceptable to the public has been costly in terms of conflict, delay and litigation and the expenditure of labor, time and money that could have been used more efficiently elsewhere (Behan 1990; Gericke and Sullivan 1994; Shands et al. 1990, Wondolleck 1988).

The controversies that have evolved over public involvement in National Forest planning have mirrored problems in other areas of policy making. Senator Hubert Humphrey, the chief sponsor of NFMA, desired a law that would ensure that National Forest planning was not left solely to the domain of technocrats in the Forest Service, but would incorporate public preferences and desires in the planning process (Humphrey 1976). Humphrey’s sentiments have been echoed by the many policy practitioners and theorists who have argued that technically-driven, value-free, objective solutions to policy problems do not exist and maintain that greater emphasis needs to be placed on public involvement in policy analysis and formation (Dryzek and Torgerson, 1993; Laird 1993; Fischer 1993: Fiorino 1990; deLeon 1988, 1995; Renn et al. 1995). While greater public involvement has been seen as one way to incorporate the values and preferences of the public into the decision making process, the challenge has been in finding techniques that systematically analyze participant perspectives.

Surveys, contingent valuation, focus groups, and multiattribute utility analysis all have been employed by policy analysts to articulate the public’s priorities among diverse goals (Fischhoff 1991; Gregory and Keeney 1994; Keeney and Raiffa 1976; Keeney et al. 1990). Surveys and contingent valuation suffer from difficulties in designing and administering the questions and interpreting the results (Keeney et al. 1990). Focus groups are often small and unrepresentative and no specific guidelines exist to elicit a systematic understanding of value-relevant information (Keeney et al. 1990). Multiattribute utility analysis is often difficult for lay people to understand (Keeney et al. 1990). An inductive, yet systematic, methodology is needed to provide information on public viewpoints, values and positions.

Q-methodology has emerged to fill this void. Q-methodology promises to lend sharper, more systematic insight into the values and preferences held by the public. Oftentimes bureaucrats, researchers and analysts prefer to deal with ‘facts’ or empirically established data and therefore avoid addressing value issues and public preferences. This can occur because subjectivity, or an individual’s personal point of view, is often thought to be difficult, if not impossible, to study with any degree of precision. Q-methodology provides researchers and analysts with "a systematic and rigorously quantitative means for examining human subjectivity" (McKeown and Thomas 1988).

This paper illustrates how Q-methodology can be used to better understand participant concerns in National Forest management. Two case studiesin National Forest management are reviewed in this paper and demonstrate how Q-methodology can (1) identify important internal and external constituencies; (2) define participant viewpoints and perceptions; (3) provide sharper insight into participant preferred management directions; (4) identify criteria that are important to participants, (5) explicitly outline areas of consensus and conflict; and (6) develop a common view toward the policy. The first case explores residents’ beliefs about ecosystem management in the Chattooga Watershed in the southern Appalachians. The second case study investigates USFS staff perceptions of National Forest planning on the Monongahela National Forest (MNF) in West Virginia. The Chattooga case employed a Q-analysis after soliciting input from the public in the form of a mail questionnaire. The MNF study used Q-method in a focus group setting. The strengths and weaknesses of each approach are evaluated. Criteria for estimating the usefulness of Q for policy analysts and researchers are explored.

What Is Q-Methodology?

Q-methodology is a technique for studying human subjectivity (Stephenson 1953; McKeown and Thomas 1988; Brown 1980; 1993; Durning 1996). Every person perceives the world differently, and Q-methodology uses these subjective viewpoints to construct typologies of different perspectives. Q-method was developed in the 1930s by William Stephenson, a British physicist-psychologist (1935a, 1935b, 1953). Stephenson was interested in using factor analysis to correlate people with the views they held to reveal the multiple points of view that could prevail in any situation (Brown 1996, 1993).

There are two discerning factors that set Q methodology apart from R methodology. The first is the subjective nature of Q methodology versus the objective nature of R methodology. The second is the nature of the correlation and clustering that occurs with Q methodology versus R methodology.

R methodology studies the relationships among objective variables (Brown 1984). For instance, R methodology would correlate the objective variables of a high income and a high degree of education with the variable of a high degree of political participation. High income in this case might be a quantifiable annual income. High education could be quantified by the exact number of years in school. Political participation might be identified as participating in the most recent election. All of these variables are objective, externally verifiable answers where "proof and refutation are at issue" (Brown, 19993: 2). The power of R methodology is in abstracting these traits and attributes from the individuals who possess them and generalizing the findings to explain the characteristics of populations. Consequently, R-methodology typically uses large samples of subjects to explore variability between cases (Rohrbaugh 1997b).

Unlike R-analysis, which is concerned with patterns across variables, Q-methodology is concerned with patterns of subjective perspectives across individuals. A Q-methodology approach would ask a group of individuals what factors they felt were associated with political participation. All of the statements contributed by the group would be subjective, that is, there would be no right or wrong answer. Some people might feel that a strong family history of political involvement might contribute to political participation. Others might feel that being a member of the Democratic Party was a factor. Still others might cite that being involved in local associations played a role in political participation. All of the people who contribute statements would each be asked to rank order the collection of statements to reflect their personal view of how the various factors relate to their beliefs about political participation. The variation in the ordering of the statements would be subjected to statistical analysis to reveal dominant patterns of belief. In this way, Q method identifies how individuals with like views perceive an issue (Durning, 1996). Likewise it reveals how individuals with different views see the issue. Q is less concerned with the ability to generalize the findings from the analysis and uses smaller, well-selected samples to analyze variability within cases (Rohrbaugh 1997b). Q analysis does not yield statistically generalizable results. Instead, the results produce an in-depth portrait of the typologies of perspectives that prevail in a given situation.

Some people incorrectly believe that Q-method is simply an inverted R-methodology or "that it is really nothing more than the application of R-method factoring technique to a transposed data matrix" (McKeown and Thomas, 1988: 47). However, this is incorrect, and it is important to clarify exactly what is being factored. As described above, Q-method implies that correlation and factoring of persons. R-method implies the correlation and factoring of traits. With R-methodology, correlation summarizes the relationships among the traits and then factor analysis denotes the clusters of traits. With Q-methodology, correlation summarizes the views among the people and then factor analysis denotes the clusters of people. In R-method, the traits are centered in columns and correlated together. For instance, income would appear in column A and years of education would appear in column B. In Q-method, the various beliefs of individuals would be centered in columns and there is no common unit of measurement. For instance column A might be Bill and he might have agreed strongly with the statement that "Being a member of the Democratic Party" and "Being a member of associations" are closely tied to political participation. In column B might be Ken and he might disagree strongly with the same statements. These statements would be correlated to reveal differing perceptions on behalf of these two individuals. The only common unit of measurement would be the individual’s self-referential viewpoint (McKeown and Thomas, 1988; Brown, 1996). Consequently, the correlation and factoring process are very different between Q and R, and as Stephenson, the founder of Q-methodology observed, "There never was a single matrix of scores to which Q and R apply" (1953: 15).

Public Involvement and Q-Methodology

Two broad principles are useful for conceptualizing public involvement with respect to Q-methodology. The first is the relationship between the policy analyst or decision-maker and the public. The second is the stage of the policy process where the information gathered from the public will be used. Together, these two parameters imply that the analyst knows the type of information that will be yielded from the exercise and how that information will be used to facilitate decision-making.

Several authors (Arstein 1969; Burke 1968; Roberts 1995) have demonstrated how to conceptualize the relationship between the policy analyst or decision-maker and the public. Roberts (1995) divided public involvement into two categories—consultation and participation. Roberts’ classification hinged on the degree to which the public was given decision-making power. Consultation included education, information sharing, and negotiation. Participation conveyed that the public was given actual decision making power. Most often, the public is involved in consultation and not participation. For instance, decision-makers are often interested in learning the public’s views, opinions and preferences, but not in giving the public power to formulate actual policy.

In addition to conceptualizing the relationship between decision-maker and the public, it also is important to understand at what stages in the policy process public involvement is most helpful. With respect to public involvement Q-methodology can be used in all phases of the policy process—intelligence, promotion, prescription, invocation, application, termination and appraisal (Lasswell 1971).

In the two case studies explored in this paper, the public was viewed in a consultative manner. The public was not invited to make the decision, but was asked to contribute information that would aid in goal clarification, trend descriptions and, in some instances, give their own analysis of conditions. In this manner, the public was seen as providing information to aid in the intelligence function. The intelligence stage of the process includes the "gathering, processing and dissemination of information for the use of all who participate in the decision process" (Lasswell 1971: 28). At this stage of the decision process, Q-methodology can aid in the clarification of public values thereby providing important information to decision-makers and the public alike. The findings from this study also contributed to the promotional stage of the policy process. The promotional stage of the process is distinguished by the dissemination of values or viewpoints. By rendering values more explicit, Q-methodology aids in promotion by allowing decision-makers and the public to advocate their views. Ideally, the results from the study will aid in the prescription of policy, but this stage of the policy process has not yet occurred. The prescription function is "characterized by the stabilization of expectations" regarding the norms that are established (Lasswell 1971:29). Q-method is useful in the prescription context because it can allow a public dialogue to take place regarding values and can then contribute to the stabilization of expectations needed to achieve prescription outcomes.

Case #1: The Chattooga Watershed:

Understanding Residents’ Priorities for Ecosystem Management

In 1992 then United States Forest Service (USFS) Chief Dale Robertson introduced ecosystem management to the agency. As part of this introduction, Robertson stated that "ecosystem management means the Forest Service will use an ecological approach to achieve multiple-use management of the National Forests and Grasslands by blending the needs of people and environmental values to sustain diverse, healthy, and productive ecosystems." Since then, the approaches and components of ecosystem management have been catalogued and defined by others, and there has been widespread agreement that socially defined goals and management objectives are important in ecosystem management (Robertson 1992; Grumbine 1994; Moote et al. 1994; Clark and Minta 1994; Williams 1994; Brunner and Clark 1995). Ecosystem management is seen as more than a technical concept and must strive to incorporate the human context if management solutions are to be accepted and understood by the public. Therefore it is important that resource managers understand the social values, attitudes and opinions that relate to the resources under question.

In keeping with this view, this study explored the beliefs and values of Chattooga watershed residents to facilitate ecosystem management in the area. The Chattooga watershed straddles the three states of North Carolina, South Carolina and Georgia. The Q-analysis performed in this study was an extension of a research project conducted by Maguire as part of the U.S. Forest Service’s Chattooga River Ecosystem Management Project. The earlier analysis by Maguire (1995) investigated the desired future conditions (DFCs) of residents in the watershed. Maguire used written responses and conducted in-person interviews with 143 residents of the Chattooga watershed. The data collected during this project were used as the basis for the Q-analysis.

Citizens often possess a diversity of interests, values and opinions that can conflict or be held inconsistently. These preferences and beliefs can be difficult to assess and the goal of the Q-analysis was to provide a systematic, inductive analysis of residents’ views. Since these attitudes and opinions shape the assumptions that underlie management prescriptions in the Chattooga watershed, these data give the USFS insight into the basis of the beliefs and perceptions of its constituents.

The Q-Sample

The statements to be ranked by the residents (the Q-sample) were drawn from 143 in-person interviews and written responses that had been conducted with residents of the Chattooga watershed between November 1993 and July 1995 by Maguire (1995). A total of 55 sample statements (Please refer to Table 1 for a list of the statements) were taken from specific comments individuals made about ecosystems and their management. Topics addressed in the interviews and written responses included even-aged and uneven-aged management, fire, roads, fish, wildlife, and siltation.

There are two basic techniques for selecting the statements to be included in the Q-sample—unstructured and structured sampling (McKeown and Thomas 1988:28-29). Unstructured sampling provides a survey of positions or perspectives with respect to the issue under investigation. As a consequence, unstructured sampling can suffer from bias in the sampling of topics. In contrast, structured samples purposely select statements to cover a range of topics and seek to avoid biases in over- or under-sampling of particular subject areas.

We employed an unstructured sampling technique for the Chattooga study, but balanced the statements to ensure that neither pro nor con viewpoints were over-represented. The statements were categorized into five issue areas: Forests, Wildlife, Roads, Water, and Recreation. These issue areas were selected based on the topics raised by the residents in their interviews and written responses. The statements in each of these areas were selected to reflect the full range of views expressed by the residents. The statements were then edited to enhance clarity and repetitious statements were eliminated. There were 18 questions about Wildlife, 16 questions about Forests, 9 questions about Water, 7 questions about Recreation, and 5 questions about Roads. The number of questions reflected the relative importance of the topics as raised by the residents in the interviews and written responses.

The Person Sample

The survey was sent to all of the individuals who had participated in the initial interviews and written responses conducted during the 1993-95 period (Maguire 1995) and in public meetings held to publicize that research. The initial list of potential respondents developed in 1993-95 was compiled from names and organizations suggested by employees of the USFS three ranger districts in the Chattooga watershed and the "key contacts" list for each of the districts. A "snowball" sampling technique (Goodman 1961) was used to verify that an adequate number of respondents had been suggested at that time. The sample was diversified according to geographic location, gender and age. The final population sample was composed mainly of residents of the Chattooga watershed, while some were from the surrounding communities outside the watershed. Local members of national organizations with interests in the Chattooga, such as the Sierra Club and Trout Unlimited, also were interviewed. As people were interviewed, they were asked to recommend additional names. They were asked specifically to recommend people whose views were different from their own. This helped to ensure that a diversity of opinion was collected.

A total of 143 surveys were sent out. A dollar bill was placed in the survey envelope to encourage the recipients to fill out the survey. Post cards were sent out as reminders at two and four week intervals after the first mailing. The estimated total cost for the survey was $366.15. A total of 87 surveys were returned, for a total response rate of 60.83%. Of 87 surveys that were returned, 19 were unusable due to omissions or incorrect markings. This left 68 surveys, for an amended response rate of 47.05%.

Administering the Q-Sorts

Q-sorts are conducted w hen a subject models his or her point of view by rank-ordering Q-sample statements along a continuum defined by a set of instructions. In this case, the Chattooga residents were asked to rank the 55 statements from strongly disagree (-4) to strongly agree (+4). In many cases where Q-sorts are performed, subjects are asked to place the statements in an inverted quasi-normal distribution. Placing the statements in the quasi-normal distribution encourages the respondents to consider the relationships among the statements more systematically. However, performing the inverted quasi-normal Q-sort is a confusing task. Since we opted to use a survey, we decided that asking the respondents to perform a Q-sort without anyone to explain the procedure in person was too complex. The response rate would have suffered and the literature on the topic indicates that it is not necessary to have the respondents perform a forced distribution of the statements (McKeown and Thomas 1988: 34-35; Brown 1971, 1985; Cottle and McKeown 1980).

The 68 responses to the 55 statements were correlated in a 68 by 68 matrix. The matrix was factor analyzed using the QMETHOD software. The initial factor loadings were determined automatically by QMETHOD, which extracted seven centroid factors. Varimax rotation was used and resulted in three identifiable factors. Of the 68 respondents, 26 clustered on factor 1, 17 on factor 2, and 10 on factor 3. Factor loadings with eigenvalues greater than 1.00 were considered significant (McKeown and Thomas 1988). Analysis of three randomly generated Q-datasets of the same dimensions as the Chattooga data showed that factors explaining less than 3-4% of the variation should be ignored. A total of five factors had eigenvalues greater than 1.00 (16.98, 15.72, 4.29, 1.97 and 1.85). The final two (1.97 and 1.85) were considered too weak to warrant further attention.

Findings: Differing Management Philosophies in the Chattooga Watershed

Factor 1

Management for Wildlife and Habitat: Minimizing Human Disturbance

Factor 1 respondents most strongly identified with minimizing human disturbance in the watershed and managing the Chattooga ecosystem for wildlife and habitat. Statements 6, 28, 26 and 33 demonstrated a concern for forest management practices that were perceived as harmful to wildlife, such as clearcutting and habitat fragmentation. (Please refer to Table 1 for a list of all the factors and the factor scores). Statements 36, 33, and 38 indicated that Factor 1 types cared about the deleterious impacts of roads and vehicular access on resources in the Chattooga and revealed that some management practices, such as forest road closures, were believed to be beneficial to wildlife.

Water quality, as indicated by a positive factor score on statements 41, 43 and 48, also emerged as a topic of concern. Siltation and pollution as a consequence of development activities were of interest and these individuals indicated a desire to manage development to avoid the degradation of water.

Factor 1 individuals disagreed strongly with the notion that clearcutting could be beneficial to wildlife or forest health, as indicated by the negative factor scores on statements 3, 19 and 25. Likewise, Factor 1 types were skeptical of the idea that multiple uses could be practiced without imposing some cost on the land (statement 10), or that the Chattooga watershed could support a variety of activities (statement 9).
 

TABLE 1: Factor Q-Sort Values for Each Statement Factors  No. Statement 1 2 3 1. Logging provides jobs and materials for houses. 0 4 2 2. We should manage for timber harvesting, but emphasize uneven management. -1 -1 3 3. Areas are made healthy through clearcutting. -4 1 -3 4. Subsequent growths are never as good as old growth because the quality of timber goes down. 1 -3 1 5. After clearcut, the forest will never return to its previous state. 2 -3 1 6. Clearcutting causes erosion into streams, hurting the trout fisheries. 4 -2 2 7. A forest with mixed forest tree species types is best for wildlife. 2 2 4 8. In the long run, selective-cutting produces more timber than one harvest every 70-80 years. 1 -2 2 9. The Chattooga watershed is large enough to support a range of activities, plant and timber types, and recreational areas, while maintaining a harvesting schedule that will help finance the other uses of the area -2 3 1 10. The use of our renewable resources has improved our lands, not degraded them. -4 1 -1 11. Wildfire has always been around, with an average 150-200 year period for the wildfire cycle in the Chattooga area. 0 0 0 12. "Letting things take care of themselves" does not include letting fires burn, because fires are largely caused by man. -1 0 0 13. Controlled burns will increase browse for wildlife, increasing hunting and wildlife viewing opportunities. -1 4 1 14. We have managed the forest to the point that ground litter produces an inferno when fire breaks out. -1 -1 0 15. There will be less oaks and hardwoods because the White Pines and Hemlocks will have taken over. This process is speeded up by the Forest Service’s clearcutting and replanting White Pines. 1 -3 3 16. Farming, not Forest Service practices, causes erosion and sedimentation -2 1 -4 17 Selective-cutting has more wildlife benefits than clearcutti ng because residual trees produce acorns and den sites. 1 -2 1 18 Overuse of clearcuts seems to have contributed to a loss of hardwoods and mast needed for wildlife. 2 -3 3 19. Clearcutting has increased hunting opportunities by providing more cover and food habitat. -3 1 -2 20. Clearcutting causes erosion into streams, hurting the trout fisheries 3 -2 2 21. Clearcuts help some wildlife species for a few years. 0 2 -1 22. People who argue that trees should not be cut to protect wildlife are not considering that many species of wildlife do not want to live where there is nothing but old trees -3 2 -2 23. Now you can hardly find or hear hoot owls, song birds, small game, whippoorwill; it is because They don’t have a habitat to live in, they can’t live in clear-cuts. 0 -4 -2 24. Pines don't provide as good a wildlife habitat as hardwoods. 1 0 2 25. Clearcutting has gotten a bad name. It is very beneficial for wildlife (big game). -4 1 -3 26. Many bird require an uninterrupted forest for nesting. 3 -1 0 27. Now the Chattooga River and Tugalo Lake have so much silt washed in because the mountains have Been cleared that you cannot find a decent place to fish. 0 -4 -1 28. Logging causes fragmentation of animal habitat. 3 -2 -1 29. Good forest management and game management go hand in hand. 0 3 4 30. Controlled burns will increase browse for wildlife, increasing hunting and wildlife viewing opportunities. -2 4 3 3.1 Hardwoods are valuable for wildlife and recreation. 4 3 4 32. Hunting, thinning the population of game, is good for the overall population. 0 2 1 33. Road closure helps wildlife. 3 0 -3 34. Overhunting is not a problem in the Chattooga Watershed. -2 1 -2 35. We should limit motorized vehicle access, but keep main roads open 1 1 -1 36. We should restrict off-road vehicles to designated areas only. 4 2 -1 37. We should restrict horses and mountain bikes to designated areas only. 2 2 1 38. No vehicles should be in or even near the river corridor because they cause too much trash and erosion. 2 -1 -3 39. Road access helps in fire control, insect and disease control, and game inventory and management. 0 3 2 40. I believe our water here is as safe to drink as anywhere in the world. -2 1 1 41. The river is getting more and more polluted. 1 -1 0 42. Water quality in Chattooga is the best it's ever been. -3 -1 -4 43. Development on private property is a threat to water quality. 2 0 -3 44. The water supply when I was a child was much greater. -1 -1 0 45. I t 45 I think our creeks or rivers are probably less polluted in this area than anywhere. -2 0 -1 46. Water is the most important natural resource. 2 2 2 47. "Our water has been short this year on account of all this clearcutting. -2 -4 -2 48. We should control development of resorts and houses because of siltation. 1 0 -2

TABLE 1: Factor Q-Sort Values for Each Statement Factors  No. Statement 1 2 3 49. Hardwoods are valuable for wildlife and recreation. 3 3 3 50. Rafting and canoeing harm the fish in the river. -3 -3 -1 51. Horseback riding may cause a little erosion, but not enough that it can’t be fixed. -3 0 -2 52. The fishing season should go back to a seasonal basis. We should have narrow windows rather than year-round fishing because there is too much pressure on fish populations (trout specifically). -1 -2 0 53. Rafting companies don't have a tremendous effect on the quality of the water -1 0 -4 54. The Chattooga is heavily silted; the trout and fish are not spawning properly. 0 -1 0 55. Fishing is poor in most places in the Chattooga watershed. -1 -2 0

Factor 2

Managing for Timber Production and Wildlife: Active Management

Factor 2 types agreed strongly with statement 1, which demonstrated a pragmatic belief in the need for jobs and materials supplied by logging. While Factor 1 indicated that development and timbering were detrimental to wildlife and habitat protection, Factor 2 perceived benefits from active human management including the need for controlled burning and hunting as management tools for wildlife, as the factor scores on statements 13, 22, and 32 illustrate.

Attitudes about clearcutting also set Factor 2 apart from Factor 1. The individuals that were catalogued in Factor 2 disagreed with the statements that portrayed clearcutting as harmful and that painted timbering practices as deleterious to wildlife and water resources, as indicated by the negative factor scores on statements 47, 27, and 23. Moreover, Factor 2 individuals emphasized the positive effects associated with clearcutting, especially the benefits to wildlife, as indicated in statements 21, 25 and 19.

Factor 3:

Traditional Use: Maintaining Access and Minimizing Impacts

Factor 3 individuals, like those in Factor 2, placed importance on the need for logging, as statement 1 indicates. However, those in Factor 3 advocated the importance of timber management techniques that emphasized mixed species, selective cutting methods and uneven-aged management (Statements 15, 2, and 8). Factor 3 individuals disagreed that clearcutting was a healthy and beneficial practice, as statements 3, 19 and 25 indicate.

Factor 3 types were concerned about vehicular access in the Chattooga, as the negative factor scores on statements 33 and 38 illustrate. Factor 3 individuals disagreed with the notion that prohibiting vehicle access produced benefits to wildlife or to habitat, such as river corridors. This could be interpreted as a desire to maintain access to forest resources.

Finally, Factor 3 individuals were concerned with a variety of human impacts on water quality. These individuals indicated a skepticism that farming was the culprit with respect to erosion and sedimentation, or that development was a threat to water quality (Statements 16 and 43). However, they did believe that rafting companies were having a deleterious impact on water resources (Statement 53).

Several consensus statements emerged from the analysis. These were statements that all three factors ranked positively. Statements 7, 31 and 46 reflect that water, hardwoods and mixed forest tree species are important to these residents of the Chattooga watershed. All three groups of respondents agree that hardwoods are especially important for wildlife and recreation. They also agree, but not as strongly, that mixed forest tree species and water resources are important. These statements suggest that there are areas of consensus that the USFS could focus on to build support for a management plan.

Implications for Management and Public Involvement

The input obtained from the residents of the Chattooga watershed can be helpful to the USFS by (1) defining different publics with which the agency can consult as it shapes its management direction; (2) identifying public viewpoints and perceptions; (3) providing sharper insight into the public’s preferred management directions; (4) identifying criteria that are important to the public, and (5) explicitly outlining areas of consensus and conflict.

Defining publics. If the USFS wants to engage in a dialogue with individuals to chart a management direction in the Chattooga watershed, the agency now knows that it needs to consult with individuals who represent each of these viewpoints. Not all of the individuals within each factor need to be consulted, since they all have similar views toward the management of the Chattooga watershed. This could be beneficial to the USFS in terms of cost savings. Small focus groups or personal visits could be utilized instead of soliciting public comments or holding hearings. However, if the USFS chooses to consult a smaller group, the agency and selected participants should communicate their findings to the broader public to keep them informed. Identifying viewpoints. These views are neither right nor wrong, for they represent subjective beliefs about ecosystem management in the Chattooga watershed. The agency, as well as the residents of the Chattooga watershed, should be interested in these findings. The feedback from the Q-analysis can substantiate or falsify pre-existing beliefs the residents and the USFS may have had about desired management in the watershed. For instance, the Factor 2 timber management types might want to know how the other groups perceive clearcutting. This could serve as a basis to begin a dialogue about appropriate timber practices on the Chattooga. Likewise, the Factor 1 wildlife and habitat individuals could find it useful to understand why the other two groups believe that the Chattooga is large enough to support a variety of uses. To contribute to social learning, developing common knowledge about the Chattooga, and cultivating continued public interest in participating in these types of analyses, it is imperative that researchers and agencies provide feedback from the Q-analysis to the original participants.

Providing management direction. The analysis gives the USFS a better understanding of what its public believes. This is an important cornerstone in ecosystem management. The three viewpoints suggest different management directions for the USFS in the Chattooga basin. Factor 1 individuals emphasize the need to minimize human disturbance in the Chattooga watershed for the benefit of wildlife and wildlife habitat. Factor 2 individuals emphasize an active, interventionist management strategy to facilitate timber production and wildlife management. Factor 3 individuals accentuate the need for maintaining access to the Chattooga’s resources, while minimizing the damaging impacts of resource use. The consensus statements provide a common foundation for building a management strategy that all three groups find acceptable.

Interestingly and reassuringly, the three management directions suggested by the three factors from the Q-analysis mirror three overarching management themes identified from a hierarchical coding analysis of the interview data from which the Q-sample were derived (Maguire 1995). The coding analysis also suggested water quality as a consensus goal around which forest management direction could coalesce. This confirmation of important themes from two different types of analysis, one qualitative and one quantitative, is an important means of validating research on subjective data, such as public values.

Identifying criteria. All of the groups of respondents identify criteria that are important to them in the context of their belief systems. These criteria can be helpful to the USFS since it needs indicators to measure its ability to meet the public’s needs. Factor 1 individuals indicate that wildlife and habitat are important criteria to them. Jobs and materials for housing are important to those in Factor 2. Likewise, jobs and materials for housing were also important to Factor 3 respondents, but these people were also interested in seeing mixed species of trees and alternative harvesting techniques. These criteria provide operational means for the USFS to measure its ability to manage for these different ends.

Outlining areas of consensus and conflict. The Q-analysis outlined areas where the different groups agreed and disagreed. Clearcutting emerged as an issue area where Factor 1 and 3 agreed, but Factor 2 disagreed. This suggests that there could be benefit to delving further into their belief systems about clearcutting. The timber management types in Factor 2 have a different set of priorities than the other two Factors. Perhaps there is some benefit to engaging the different groups in a dialogue to better understand these differences. Factor 1 and Factor 2 also disagreed about the amount of harm inflicted by timbering practices. Factor 1 and 3 disagreed over the potential for damage to water resources and the need to manage development in the watershed. Likewise, they disagreed over the need to close roads.

There could be benefits to investigating further the source of these beliefs. Studies could be designed jointly to resolve disputes about "facts" or beliefs. Once the areas of conflict are outlined, the USFS can then begin address them. Points of misinformation can be clarified. Core beliefs can be identified. The various groups can agree to disagree about certain points. However, until these conflicts are laid bare, they do not stand a chance of being addressed. Likewise, the points on which there is consensus may be of interest to the USFS and the residents. The consensus statements may reflect previously unknown common beliefs. When these areas are made explicit, the USFS has a new set of values with which to promote a unifying management concept.

Case #2: Monongahela National Forest Planning Process:

Understanding Priorities for Planning Among MNF Staff

According to the National Forest Management Act (NFMA) every National Forest is mandated to conduct a planning process every 10-15 years. The Monongahela National Forest (MNF) completed its first generation land management plan in 1986. Consequently, the resource managers on the MNF are now turning their sights to the next planning process.

One of the reasons public involvement in National Forest management has often fallen short of desired outcomes has been a gap in the expectations between the public and USFS staff (Steelman 1996; 1998). The public has different perceptions of what constitutes public involvement than do the bureaucrats who are in charge of implementing the public input mandate. Q-methodology can be seen as a tool to make more explicit the expectations and beliefs held by the public with respect to planning and how the public should be involved in planning.

Instead of turning to the external public, this exercise focused on the internal constituencies-- the staff and employees on the MNF. A Q-analysis was proposed to give the planning staff on the MNF some insight into what its own agency was thinking with regard to the upcoming planning process. Oftentimes, emphasis is placed on the need to solicit views from the external publics while internal ‘publics’ are overlooked (Roberts 1996). A set of questions regarding the upcoming MNF planning process was put together with the help of the MNF Forest Planner and a group of MNF planning participants. These questions were to help make explicit the way that the employees within the agency viewed the planning process and the role of the public in that process. The five questions selected focused on (1) the usefulness of the plan; (2) the planning process itself; (3) agency perceptions of public involvement; (4) the usefulness of public involvement; and (5) how the public should be involved. The questions, analysis and results were presented over a two-day workshop at the headquarters of the MNF in Elkins, WV. Fifteen participants who represented different facets of the MNF staff graciously donated their time to take part in the exercise.

The Person Sample

The person sample for the study was convened as a focus group. The focus group was selected based on their diversified backgrounds and their interest in participating in the two-day workshop during which the Q-method was used. McKeown and Thomas (1988: 36) indicate that the person sample can be constructed using either theoretical or pragmatic considerations. Theoretical considerations mean that people are chosen because of their special relevance to the goals of the study. Pragmatic considerations dictate that those who are most readily available will participate.

A combination of theoretical and pragmatic considerations was used in constructing the person sample for the MNF case study. The forest planner screened the individuals who wished to participate to ensure that different areas of expertise were represented. Those who ended up participating did so on account of their availability. Fifteen employees participated, eight women, seven men. Different geographic regions within the MNF were also represented. Areas of expertise included a recreation planner, a wildlife biologist, GIS technicians, clerical staff, a timber management professional, a district ranger, a civil engineer, a NEPA officer, a botanist, a timber inventory specialist, a GIS coordinator, a visitor center planner, and a land management planning specialist. This broad geographical, gender and professional range ensured that a variety of viewpoints would be presented during the exercise.

The Q-Sample

The statements for the five Q-samples came from discussions held in the focus group. Different Q-samples were created for each question, since the questions addressed different issue areas. Eleven statements provided the Q-sample for question 1: What are the most/least useful aspects of the planning process? Twenty-nine statements were used in the Q-sample from question 2: What aspects of the plan work or don’t work? Twenty-six statements formed the basis of the Q-sample for question 3: What does the public feel the plan does or does not address? Eighteen statements emerged from question 4: What is most or least useful with respect to public involvement? Twenty-one statements provided the Q-sort from question 5: How should the public be involved? All of the statements listed in the Q-samples were the direct comments of the individuals participating in the focus groups. As such, some of the statements may not be clear or make sense to others who were not part of the original exercise.

Administering the Q-Sorts

After the Q-samples had been constructed, the focus group participants were asked to place the statements into a Q-sort by putting them into a normal distribution. This exercise was repeated five times with each question providing a different Q-sample. The Q-sorts were analyzed using the same methods as the Chattooga data. Four factors emerged as significant for question 1, two for questions 2 and 5, and three for questions 3 and 4.

Findings

Question 1: What are the most and least useful aspects of the planning process?

There were four significant viewpoints with respect to question one. Table 2 lists the statements and the four factor scores for question 1. Factor 1 individuals were distinguished by their belief in the need for greater compromise and negotiation in the planning process, as indicated by their responses to statements 2 and 3. Factor 2 placed an emphasis on communication with the public. In statements 1 and 10 Factor 2 respondents asserted that there were inconsistent expectations between internal and external publics and that information needed to be more comprehensible to the public. These respondents also revealed that negotiation and compromise were not important to a useful planning process (statements 2 and 3). Factor 3 respondents placed an emphasis on the need for information in the planning process. Information needs to be comprehensible to the public in addition to being detailed and analytical (statements 9 and 10). Factor 4 individuals wanted planning to be a more dynamic process (statement 5). They also indicated that external factors, such as knowing the MNF leadership’s expectations, were important (statement 8). Factor 4 respondents indicated that information was not particularly important in the planning process (statements 9 and 10).

Three consensus statements emerged from the analysis. The group believed that the security of their positions was not related to effective planning (statement 4), nor did they feel very strongly about the need to have the appropriate budget (statement 6) or a realistic workload to accomplish their planning tasks (statement 7). The 1990s have been a time of great upheaval in the USFS, with many reductions in forces and threats to job security. While these issues were raised during the discussion to formulate the Q-statements, they did not take priority when the group ranked them in their Q-sorts.

In summary, four factors emerged as the most or least useful aspects of the planning process. Factor 1 emphasized the usefulness of negotiation and compromise. Factor 2 indicated that better communication with the public was helpful. Factor 3 placed a priority on the need for good information in the process. Factor 4 stated that direction from the leadership within the MNF and making the process more dynamic were most useful.
 

Table 2: Factor Scores for Question 1 What is most/least useful in the planning process? Factors 1 2 3 4  1. Inconsistent expectations on revision between internal and external publics. 0 3 –3 1 2. Need negotiation and respectful listening among teams. 2 –2 0 0 3. Greater emphasis on compromise and negotiation. 3 –3 –1 0 4. Positions need to be secure in order to protect resource appropriately. –3 –1 –2 -3 5. Planning needs to be more dynamic and less static. 0 1 1 3 6. Need appropriate budget. 0 0 1 1 7. Realistic workload. –1 0 0 0 8. Need to know leadership’s expectations. –2 -1 -1 2 9. Need detailed, analytical information on forest. 1 0 2 -1 10. Information needs to be comprehensible to public. 1 2 3 -1 11. Information needs to be compatible 0 3 -3 1

Question 2: What aspects of the plan work or don’t work?

Two factors were significant with respect to question 2 (Table A in the Appendix). Factor 1 and 2 individuals both agreed with statements 1and 3 that one of the most helpful aspects of the plan was that it provided a general, as well as specific, foundation and direction for action on the Forest. Additionally, to achieve both broader and specific objectives, both Factor 1 and 2 respondents agreed with statement 7 that standards and guidelines were helpful to give them direction.

Factor 1 was distinguished from Factor 2 by generally being more satisfied with what the plan did. For instance Factor 1 individuals disagreed that the plan was too complicated to change if there was a problem with it, as indicated by the response to statement 11. Factor 1 indicated that the plan did a better job of balancing technological and social concerns than Factor 2 types, as indicated in statement 23. Factor 1 individuals responded more positively to statements 28 and 29 and were more likely to be satisfied with the delineated management areas than Factor 2 individuals. Factor 2 individuals felt more strongly that the plan needed to be written in understandable language (statement 20).

In summary, both Factor 1 and Factor 2 individuals like the fact that the plan provides general as well as specific direction. However, Factor 1 individuals seem more satisfied with what the plan does than Factor 2 individuals. Factor 2 individuals find more aspects that don’t work or could perhaps work better.

Question 3: What does the public feel the plan does or does not address?

The group was asked to comment on what they believed the public thought the plan did or did not address. The goal was to elicit agency beliefs about the public and the plan. There was agreement regarding what the plan addressed. However, three factors emerged from the analysis based on issues that the plan did not address (Table 4). Factor 1 typified the public as feeling that plan does not speak to threatened and endangered species (statement 4), old growth issues (statement 5), and biodiversity issues (statement 8) and that it doesn’t incorporate new information or changed conditions (statement 21). Factor 2 indicated that the public feels the plan does not address where and when people can hunt (statement 2), why they can’t collect firewood in certain areas (statement 3), or the need for motorized use (statement 13). Factor 3 individuals indicated that the public sees the plan as not addressing the need for access and motorized use (statements 13 and 14) and that the plan is inconsistent, impersonal and inflexible(statements 24, 25, and 26). They are also more confident that new information and changing conditions are being incorporated into the plan.

All groups agreed that the public perceived that the plan addressed where timber could be cut, the limits on commodity outputs, and guidelines for special areas (statements 1, 15 and 12). They also saw the plan as allowing multiple uses to occur in the Forest and describing the desired future conditions in the Forest (statements 11 and 17).

Question 4: What is the least or most useful with respect to public involvement?

Three factors emerged regarding what the group felt was most or least helpful in trying to involve the public in forest planning (Table B in Appendix). Factor 1 individuals felt that the public was most helpful when they were willing to be flexible and engage in a dialogue with agency officials (as indicated by statement 12). They liked it when the public presented them with new information (statement 16), but were less concerned with being presented site-specific information (statement 3) than the other two Factors. They were also less concerned about personal attacks than the other two Factors, as the factor score on statement 2 illustrates.
 

Table 4: Factor Scores for Question 3 What does the Public feel the plan does or does not address? Factors 1 2 3  1. Where timber can be cut and can’t be cut. 4 3 3 2. When to hunt, where to hunt. 2 -4 3 3. Collect firewood in specific areas. 1 -3 0 4. T&E species. -4 -3 0 5. Old Growth issues. -3 -2 1 6. The public isn’t aware of plan. 0 -1 -1 7. The public doesn’t know who USFS is . . . lack of identity. -1 2 -1  8. Biodiversity. -3 -1 1 9. Economic development. -2 0 1 10. Use for ORV’s. -2 0 1 11. Does describe desired future conditions. 4 4 4 12. Like guidelines/directions for special areas. 2 3 3 13. Need for motorized use. -1 -4 -4 14. Need for access. 1 -2 -4 15. Does set limits on commodity output. 3 3 2 16. Likes continued use/existing use to continue. 3 2 2 17. Allows multiple land use. 2 4 4 18. Like knowing there is a Plan. 1 0 2 19. Doesn’t go far enough. 0 0 0 20. Plan is overwhelming. -1 -1 -2 21. Changed conditions and new information. -4 2 -2 22. Recreation opportunities. -3 -2 0 23. Partnerships/coops. 0 1 -2 24. Consistent implementation. -2 -1 -3 25. Personal document. 0 1 -2 26. Plan is not black and white-- flexible. 3 -3 -3

Factor 2 individuals found it helpful when the public aired their complaints (statement 1), but didn’t like it when the public launched personal attacks (statement 2). They also liked it when the public could give site-specific information or expressed its preferences for desired future conditions (statements 3 and 4). This group found that positive feedback from the public was not as helpful or necessary to them as it was for the other Factors (statement 7). Likewise this Factor did not find it particularly useful when groups worked out problems for themselves (statement 5).

Factor 3 types didn’t like hearing negative comments from the public. They did not like it when the public aired their personal gripes or when they waged personal attacks (statements 1 and 2). This group was mostly interested in getting site-specific information from the public (statement 3). It was especially helpful to Factor 3 individuals when the public could offer potential solutions (statement 10).

Consensus statements included agreement on the fact that involving the public was very time-consuming (statement 13). All factors agreed that not getting feedback from the public was not very helpful (statement 15). On the positive side, all three groups liked it when the public could articulate the desired future conditions it wanted to see in the MNF and when the public offered potential solutions to management problems (statements 4 and 10).

In summary, three different versions of what constituted helpful public involvement emerged from the analysis. Factor 1 individuals found a public that was willing to be flexible and that could provide them with information most helpful. Factor 2 liked it when citizens could articulate their complaints and provide site-specific information. Factor 3 did not like it when the public was negative and favored getting site-specific information and solutions from them.

Question 5: How should the public be involved?

Two factors emerged from this analysis (Table C in Appendix). Factor 1 placed greater emphasis on individual and personalized communication. Their preference would be to talk to everyone one on one. In contrast, Factor 2 favored being able to communicate with larger audiences. They placed emphasis on public comments and public meetings as the most effective way to involve the public. Factor 1 placed importance on one on one meetings with the public, distilling the message to reach the public, weekend gatherings, and getting information to interest groups (statements 3, 2, 8, 21 and 18). Factor 2 emphasized public meetings and written public comments (statements 1 and 20) as the favored means to involve the public.

Consensus statements indicated that both groups agreed that many methods of involving the public were not helpful, such as working through Wal-Mart, local festivals, or sporting goods stores (statements 12, 16, 17). Newsletters, attending meetings, targeting specific groups and utilizing media sources such as the television, newspapers or the radio received lukewarm enthusiasm from both groups (statements 13, 5, 7, and 11). Both factors thought that working one on one with the public (statement 3) was important, but Factor 1 placed a greater emphasis on this than Factor 2. Both factors again agreed that tailoring the message was important (statement 19), but Factor 2 felt more strongly about that than Factor 1.

Implications for Management and Public Involvement

The input obtained from the participants in the MNF focus group can be helpful by

(1) identifying multiple constituencies within the agency; (2) defining internal viewpoints and perceptions; (3) providing direction to the planner and planning staff based on internal views; (4) explicitly outlining areas of consensus and conflict, (5) developing a common view toward planning and public involvement.

Identifying Internal Constituencies—One of the insights that evolved from this analysis is that employees on the MNF do not look at planning and public involvement issues in the same way. Many different types of people work for the MNF. Oftentimes we assume that because people are employed by the same agency they have similar outlooks on the issues with which they work. The data from the Q-analysis indicates that this is not true. Multiple viewpoints emerged on each of the questions posed. Individuals who work in the same professional area do not necessarily hold the same views on these subjects. Therefore, as the planning staff on the MNF proceeds with the development of a new plan, it must keep in mind that the constituencies within the agency could be as complex as the public external to the agency. This cautions the USFS, as well as other agencies or organizations, to pay attention to its own constituents and do its internal work before going to its external publics. Interestingly, the interview portion of the Chattooga study, which included group interviews of Forest Service employees, produced a similar result: the views of agency employees are nearly as diverse as those of the general public (Maguire 1995).

Defining Internal Viewpoints and Perceptions and Providing Direction to the Planner and Planning Staff--- Q-methodology identified very specific viewpoints within each of the issue areas raised by the questions. For instance, with respect to the least and most useful aspects of the planning process, four distinct perspectives emerged. Some employees placed a greater emphasis on compromise and negotiation in the process than others. Establishing good communication with the public was felt to be more important by some. Others placed an emphasis on the need for information in the planning process. However, when asked which aspects of the plan do or don’t work, there were only two distinct viewpoints- some were more content with the plan than others. For the most part, there was a great deal of consensus on this issue. This information lets the planner know that most employees see the plan as a useful tool, but that they have differing views about which aspects of the planning process need to be improved. Therefore, the planner should not spend his time worrying about or justifying the need for the plan, but should rather concentrate on assuring that good information is used in the process, making the planning process more dynamic, and developing negotiation and communications skills within the agency.

Likewise with respect to public involvement, the employees held three distinct perceptions of the public’s view of planning. These three views may reflect the different types of external publics with which these employees come into contact. All three groups identified areas where the current plan is perceived as weak, thereby suggesting future direction for the next plan. These areas include old growth, threatened and endangered species, and biodiversity (Factor 1), providing better direction with respect to hunting and collecting firewood (Factor 2), and tackling the issue of access and motorized use, not to mention making the plan more user friendly for the public (Factor 3).

Question 4 asked what the most and least useful aspects of public involvement were. The analysis for this question revealed that different groups within the agency have different preferred ways of interacting with the public. Some look to the public to be a partner and engage in dialogue and provide new information, others are more sensitive to criticism and only want specific information from the public, others want the publics’ criticisms so that they can react to them. Again, these preferences can be interpreted based on the contact that these employees have with the public. Those who have been unduly criticized by the public are weary of their negativism. Those who want to hear the public’s complaints have been trained in public involvement techniques and are prepared for potentially harsh comments. These findings suggest that the planner and other officers within the USFS could better prepare their employees to deal with the public or channel public comments to those who are most capable of dealing with them.

Finally, question 5 asked what were the most effective ways of involving the public. The viewpoints that evolved from the analysis indicated two distinct methods for involving the public. The first one could be characterized as a two-way communication philosophy. This group wanted to engage in one-on-one dialogue in a variety of settings. The other view looked at public involvement more as a one-way communication effort. This group focused on public meetings and comments. While this method can be seen as less labor intensive and time consuming, it allows for only limited two-way interaction. Again, this suggests that there may be differing needs within the MNF for communicating with the public. Those in the first category may need more direct involvement with the public to be effective and may also have the time to commit to such efforts. Those in the second category may have different needs or be too constrained by time to undertake more labor- and time-intensive one-on-one communication efforts. Therefore, the MNF may wish to evaluate how and why it needs to communicate with the public and which strategies are most effectively employed to meet its desired goals.

Explicitly Outlining Areas of Consensus and Conflict—Engaging in this exercise clarified areas of consensus and conflict where only vague notions existed before. The MNF planner was pleasantly surprised to discover that the employees almost unanimously found the plan to be helpful. Likewise, the planner was surprised that there was not more conflict among viewpoints. While the Q-analysis highlighted differences in perspective, the method also identified areas of overlap. These data can help the planner and the planning staff chart a course in the next planning process that emphasizes areas of consensus. Likewise, the staff has been alerted to areas of conflict and can now begin to address those areas proactively.

Developing a Common View-- Since the exercise was conducted over a two-day period it allowed for sustained contact, dialogue and exchange among the participants. The employees were able to participate in the analysis and contribute to the interpretation of the feedback when the results were presented to them. As a consequence of the group exercise, a common view emerged from the group—that planning was a valuable tool and that the public needed to be involved in it. While these might seem like broad assertions, they were an unexpected outcome. Since planning has been so contentious, it was reassuring that there were common attitudes among the group.

Trade-offs with Surveys and Focus Groups

At the beginning of this paper it was suggested that Q-methodology would be used in a consultative manner with the public. The goal of this study was to consult with various stakeholders in the Chattooga and on the MNF about their views on ecosystem management and National Forest planning. Two methods of consulting with the public were utilized—a written survey and a focus group. At the outset of this paper, Q-method was also posited to aid in the intelligence, promotion and prescription functions of policy analysis. Specifically, within these stages it was expected to aid in goal clarification, trend description and the analysis of conditions for the given issue area. This section of the paper evaluates the two methods of obtaining responses (Table 4).

Both the written survey and the focus group were useful in providing information to the decision-makers. However the two methods differed in their capacity to process and disseminate the information to other participants once it had been gathered. The survey and the focus group were helpful in gathering information about the Chattooga watershed and the MNF. The nature of the survey limited the ability of the Chattooga residents to discuss, process and disseminate the information to other residents in the Chattooga. The survey respondents’ contact with the decision-maker was limited to filling out the questionnaire. In contrast, the MNF focus group was able to discuss and process the information during the focus group sessions. Once the Q-sorts were completed, the focus group was given the results of the analysis and asked to comment on them to determine if the outcomes accorded with their own beliefs. In this manner, discrepancies in the data and counter-intuitive findings were discussed and resolved. The ability to continue to develop a dialogue around the issues of planning and public involvement provided an additional check on the validity of the data gathered. Dissemination of the information was immediate. In contrast, the information gathered with the Chattooga survey relied on the researchers and the USFS to report the results back to the respondents.

The promotion stage of the policy process focuses on the dissemination of values and viewpoints. Since Q-method renders viewpoints more explicit, the technique can aid in the promotion of different sets of values. Once again, the focus group method is more useful in this context than the survey. Since the participants in the MNF focus group were given the results and then had the opportunity to discuss the viewpoints that emerged, they were given the fodder for additional dialogues on the topics. The views and preferences of different groups were now more explicit and could serve as the basis of future discussions and debates on the topics of planning and public involvement. However, the extent to which the participants actually disseminated these viewpoints is difficult to assess. The survey respondents in the Chattooga have not been given their results yet, nor have plans been made to bring them together in a forum to discuss the results. Consequently, the opportunities to engage in promotional activities are limited. The promotion of the values and viewpoints from the Chattooga study will depend heavily on intervention from the researchers and the USFS.

The prescription stage involves the stabilization of expectations regarding the norms that are established. The potential for Q-method to aid in prescription is questionable, since it involves such a small number of participants. The survey method limits prescription because the results are not easily communicated back to the respondents and the opportunity for creating a wider dialogue to establish norms is constrained. The promotion of the values and viewpoints
 

Table 4: Degree of Participant Involvement in Survey and Focus Group Approaches
	Chattooga Survey	MNF Focus Group
Intelligence •Gathering information •Discussing/processing information     •Dissemination of information	Very Involved  Less Involved – dependent on researchers and USFS Less Involved-- dependent on researchers and USFS	Very Involved Very Involved     Very Involved
Promotion •Dissemination of values and viewpoints	Less Involved—dependent on researchers and USFS	Potentially Involved—difficult to assess the degree to which the participants will disseminate information
Prescription •Stabilization of expectations	Less Involved	Potentially Involved— Stabilization of expectations will depend on how widely views have been disseminated
Goal Clarification	Very Involved	Very Involved
Trend Description	Very Involved	Very Involved
Analysis of Conditions	Very Involved	Very Involved

from the Chattooga study will depend heavily on intervention from the researchers and the USFS. Since the MNF participants engaged in a dialogue about their views, they have the potential to begin the process of establishing norms. However, there were only 15 participants, so the likelihood of these norms establishing themselves among the broader internal public of the MNF is questionable.

The participants in both the survey and the focus group provided goal clarification, trend description and the analysis of conditions. A clear picture of the views of the Chattooga residents’ preferences emerged from the Q-analysis. The data provided insight into the residents’ understanding of ecosystems and their priorities for management. Likewise, the MNF focus group clarified their preferences for planning and public involvement in land management, while describing some of the past trends and conditions that underlie these viewpoints.

Using Q-Methodology:

Criteria to Consider When Selecting a Survey or Focus Group

The strength of the survey-based Q-analysis is that it can be a relatively cost-effective way to collect subjective information through an inductive method. The weakness is that it loses some of its contextual benefits when isolated from the population from which the data is collected. The MNF example is strong because it gives a contextually based, inductive analysis of the views of the Forest Service staff. The drawback is that it is very cost- and labor-intensive to execute. These insights suggest that there are criteria that decision-makers should consider when contemplating a Q-method approach (Table 5).

The problem with fixed response surveys is that we assume that everyone who gives the same response to a question has interpreted it the same way, and that the meaning attached to the responses are those intended by the researcher sending out the survey. The methodology presumes that communication is perfect—that the words, phrases and language mean the same things to the same people. In the case of the Chattooga survey, even though the statements were taken directly from comments the Chattooga residents made, there is still some question regarding the purity of language and communication. Indeed, several of the Chattooga respondents commented that the survey contained unclear wording or was confusing. Since clarifications could not be made at the time the surveys were filled out, there will always be questions regarding the interpretations of the statements and the responses. With the MNF focus group, there are fewer questions with respect to interpretation of the data since the respondents generated the statements collectively and could discuss their meaning before performing their Q-sorts.

When deciding which type of method to choose, the decision-maker may wish to consider how important it is to promote direct dialogue among the group to be sampled. The focus group also provides face to face contact in a setting where dialogue and understanding are promoted. A survey may be appropriate in cases where bringing sides together would result in great conflict. There are additional ways of communicating the results to the participants and promoting direct dialogue, and understanding may be less important than obtaining the needed feedback from the consulted public. One can imagine cases where direct contact would be prohibited by distance or geography.
 

Table 5: Criteria to Consider When Selecting Survey or Focus Group Approaches
	Survey	Focus Group
Purity of language and communication	Greater question of interpretation	Less question of interpretation
Face-to-face contact	Less—but can be an asset in some contexts	Greater
Interpretive Role for Participants	Less	Greater
Interpretive Role for Analyst	Greater	Less
Time Cost	Potentially Less	Potentially Greater
Labor Cost	Potentially Less	Potentially Greater
Money Cost	Potentially Less	Potentially Greater
Flexibility	Depends	Depends

The survey results require that the researcher play a greater role divining the points of view from the respondents. Absent an audience with whom the researcher can consult readily, the analyst must rely on her own abilities to translate the meanings and intentions of the respondents. In contrast, the focus group creates a greater interpretive role for the participants in the process. The focus group is encouraged to reconcile its own beliefs with the views that emerge from the analysis, thereby validating the findings. This does not imply that the analyst is removed from the process. The statistical software still needs to be run and the researcher will provide her own interpretation of the results. Since the researcher has played a greater role in the discussions that have taken place to elicit the statements and catalogue them into the Q-sorts, she stands a better chance of understanding the context from which the results emerge. If she gets it wrong, she always has the group to verify the findings.

As has already been mentioned, time, labor and money costs need to be considered when deciding what type of method is most appropriate. The Chattooga survey was administered to 143 people with 87 responding, while the MNF focus involved 15 people. The survey cost approximately $366.15. The MNF focus group was significantly more resource and labor intensive, since it included taking 16 people away from their work and compensation for travel for the researcher. Neither of the examples included the cost for data analysis, although the costs are similar for both formats. However, a point of clarification should be made. The question of cost effectiveness becomes moot if the source of the statements is not credible. The statements for the Chattooga survey were taken from a previous research effort. Had these statements not been available, the cost for the survey would have increased. Therefore the survey option can be seen as more cost effective, provided a credible source of Q-statements exists. An attempt to minimize costs with a survey could result in the creation of meaningless data and misguided policy suggestions if the statements are not valid indicators of the individuals’ preferences and positions.

The comparative advantage of the respective methods is dependent on what aspects of the study need to be flexible. Once the survey has been sent out, the researcher will not have much flexibility to change the questions or refine their meaning. However, using a survey allows greater flexibility in terms of distance, time and resources expended. The focus group is much more flexible in terms of amending the questions or statements to make them comprehensible to the group, but is constrained in terms of setting, time of implementation, and the availability of participants.

Conclusions

In the context of the research in this paper, Q-method was proposed as a tool for consulting the public, for collecting data for the intelligence function of the policy process and for building on the promotion and prescriptive stages of the policy process. Experience using Q methodology with a written survey and focus group suggests that there are strengths and weaknesses to each approach. The focus group involves the participants in the study to a greater degree than a survey. As a consequence, the data that emerges from a focus group tends to have fewer language and communication problems. There is also greater face-to-face contact in a focus group among the participants and the researcher. This facilitates a greater degree of processing, discussing and dissemination of information than with the survey approach. The focus group can be more labor, time and money intensive than the survey. The survey can be used in situations where geography or distance prohibits face-to-face contact.

National Forest management, as well as other areas of policy, could benefit from making greater use of Q-methodology. The method’s potential for lending sharper, more systematic insight into the values and preferences held by the public could contribute to better problem identification and definition, estimation and specification of policy options, and selection, implementation and evaluation of policies.

Arnstein, Sherry R. (1969). A Ladder of Citizen Participation. Journal of the American Institute of Planners, 35(4): 216-224.

Behan, R.W. (1990). The RPA/NFMA: Solution to a Nonexistent Problem. Journal of Forestry 88(5): 20-25.

Brown, Steven R. (1996). Q Methodology and Qualitative Research. Qualitative Health Research, l6(4).

Brown, Steven R. (1993). A Primer on Q Methodology. Operant Subjectivity, 16 (3/4).

Brown, Steven R. (1985). Comments on the Search for Structure. Political Methodology. 11: 109-117.

Brown, Steven R. (1984). Q Methodology. In Directions in Social Science Research Methods, William D. Berry and Michael S. Lewis-Beck, eds. Beverly Hills, CA: Sage Publications.

Brown, Steven R. (1980). Political Subjectivity: Applications of Q-Methodology in Political Science. New Haven: Yale University Press.

Brown, Steven R. (1971). The Forced-free Distinction in Q-technique. Journal of Educational Measurement. 8: 283-287.

Brunner, Ronald D. and Tim W. Clark. (1996). Ecosystem Management: A Pragmatic Approach. Paper presented tat the 14^th Policy Sciences Annual Institute, Yale Law School, New Haven, CT. October 27-29.

Burke, Edmund M. (1968). Citizen Participation Strategies. Journal of the American Institute of Planners, 34(5): 287-294.

Clark, Tim and Steve Minta. (1994). Greater Yellowstone’s Future. Moose, Wyoming: Homestead Publishing.

Cottle, C.E. and B. F. McKeown. (1981). The Forced Free Distinction in Q-technique: A Note on Unused Categories in the Q-sort Continuum. Operant Subjectivity. 3: 58-63.

DeLeon, Peter. (1988). Advice and Consent: The Development of the Policy Sciences. New York: Russell Sage Foundation.

DeLeon, Peter. (1995). Democratic Values and the Policy Sciences. American Journal of Political Science. 39(4):886-905.

Durning, Dan. (1996). The Transition from Traditional to Post-Positivist Policy Analysis: A Role for Q Methodology. Paper presented at the Eighteenth Annual Research Conference of the Association of Public Policy Analysis and Management. Pittsburgh, PA.

Dryzek, John S. and Douglas Torgerson. (1993). Democracy and the Policy Sciences: A Progress Report. Policy Sciences 26: 127-137.

Fiorino, Daniel J. (1990). Citizen Participation and Environmental Risk: A Survey of Institutional Mechanisms. Science, Technology and Human Values. 15(2): 226-243.

Fischhoff, Baruch. (1991). Value Elicitation: Is There Anything in There?. American Psychologist. 46(8): 835-847.

Fischer, Frank. (1993). Citizen Participation and the Democratization of Policy Expertise: From Theoretical Inquiry to Practical Cases. Policy Sciences. 26: 165-187.

Gericke, K.L. and J. Sullivan. (1994). Public Participation and Appeals of Forest Service Plans—An Empirical Examination Society and Natural Resources, 7: 125-135.

Gippert, Michael J. and Vincent L. DeWitte. (1996). The Nature of Land and Resource Management Planning Under the National Forest Management Act. The Environmental Lawyer. 3(1): 149-207.

Goodman, L.A. (1961). Snowball Sampling. Annals of Mathematical Statistics. 32: 148-170.

Gregory, Robin and Ralph L. Keeney. (1994). Creating Policy Alternatives Using Stakeholder Values. Management Science, 40(8): 1035-1048.

Grumbine, Edward. (1994). What Is Ecosystem Management? Conservation Biology, 8(1): 27-38.

Humphrey, Hubert H. (1976). Remarks on the Floor, Congressional Record: S2938, March 5.

Keeney, Ralph L. Detlof von Winterfeldt and Thomas Eppel. (1990). Eliciting Public Values for Complex Policy Decisions. Management Science. 36(9): 1011-1030.

Keeney, R. L. and H. Raiffa. (1976). Decisions with Multiple Objectives: Preferences and Value Tradeoffs, New York: Wiley.

Laird, Frank N. (1993). Participatory Analysis, Democracy, and Technological Decision Making. Science, Technology, and Human Values. 18(3): 341-361.

Lasswell, Harold D. (1971). A Pre-View of the Policy Sciences. New York: American Elsevier.

Maguire, Lynn A. (1995). Desired Future Conditions for the Chattooga Watershed: A Qualitative Analysis of Diverse Public Opinions. Report to the Southeast Forest Experiment Station, USDA Forest Service, October 30.

McKeown, Bruce and Dan Thomas. (1988). Q Methodology. London: Sage Publications.

Moote, Margaret A., Sabrina Burke, Hanna J. Cortner and Mary G. Wallace. 1994. "Principles of Ecosystem Management." WRRC: College of Agriculture, University of Arizona. January 1994.

Renn, Ortwin, Thomas Webler and Peter Wiedemann. (1995). Fairness and Competence in Citizen Participation: Evaluating Models for Environmental Discourse. Boston: Kluwer Academic Publishers.

Roberts, Richard. (1996). Public Involvement: From Consultation to Participation. In Environmental and Social Impact Assessment, Frank Vanclay and Daniel A. Bronstein, eds. New York: John Wiley & Sons: 221-248.

Robertson, F. D. (1992). Ecosystem Management of the National Forests and Grasslands. Memo 1330-1, and attachments, to Regional Foresters and Station Directors from the Chief of the Forest Service, Washington, DC. June.

Rohrbaugh, Michael. (1997a). Across the Great Divide: Q vs. R in Psychotherapy Research. Department of Psychology, University of Arizona. Tucson, AZ.

Rohrbaugh, Michael. (1997b). How Crucial is Q-Sorting to Q-Methodology? A Comparison of Forced-Sort, Free-Sort and No-Sort Formats. Department of Psychology, University of Arizona. Tucson, AZ.

Rohrbaugh, Michael. (1997c). Correlating People Over Items: Applications of Q-Methodology in Psychotherapy Research. Paper presented at the North American Society for Psychotherapy Research. December 1997. Department of Psychology, University of Arizona. Tucson, AZ.

Shands, W.E. V.A. Sample and D.C. LeMaster. (1990). National Forest Planning: Searching for a Common Vision, Vol. 2. In Critique of Land Management Planning, USDA Forest Service. Washington, D.C.

Steelman, Toddi A. (1999). The Public Comment Process: What Does the Public Contribute to National Forest Management? Forthcoming in the Journal of Forestry.

Steelman, Toddi A. (1996). Democratic Theory, Public Participation and National Forest Management: A Case Study of the Monongahela National Forest. Ph.D. dissertation. Duke University. Durham, N.C.

Stephenson, William. (1935a). Techniques of Factor Analysis. Nature. 136 (August 24).

Stephenson, William. (1935b). Correlating Persons Instead of Tests. Character and Personality. 4 (September).

Stephenson, William. (1953). The Study of Behavior: Q-technique and Its Methodology. Chicago: University of Chicago Press.

United States Department of Agriculture. (1997). Office of General Counsel, Natural Resources Division. "LRMP/NFMA Forest Planning Litigation Scorecard". January 22, 1997.

Williams, Gerald W. 1994. "Ecosystem Management: How Did We Get Here?". Paper presented at the Society of American Foresters. Indianapolis, IN. February 22, 1994.

Wondolleck, Julia. (1988). Public Land Conflict and Resolution: Managing National Forest Disputes. New York: Plenum Press.
 

Table A: Factor Scores for Question 2: What aspects of the plan work or don’t work? Factor 1 2 1. Provides foundation/direction for action. 4 4 2. Provides framework for timber management. 3 2 3. Establishes area objectives.....provides standards/guidelines. 4 3 4. Flexible framework. 1 3 5. Involves the public. 1 3 6. Takes an interdisciplinary approach. 1 1 7. Creates standards and guidelines.....don’t have to start from scratch. 3 4 8. Quantifying works. 1 2 9. Subjective aspects don’t work. -2 0 10. Takes too long. -2 0 11. Too complicated to change if problem. -4 1 12. Lacks forest-wide perspective. 0 -1 13. Did not address different scales of landscape. -1 -2 14. Overly influenced by politics. -3 -4 15. Didn’t factor in human influence.....social factors. -3 -1 16. OAs conflict with boundaries from other areas. -3 -3 17. Terminology is outdated. 0 0 18. Locally issue driven. 0 -2 19. Inappropriate Land Use. -2 -4 20. Needs to be written in real language. -1 2 21. Need to use up to date technology. 0 0 22. Technology needs to be understandable to public. 0 1 23. Strike better balance between technological and social concerns. 2 -3 24. Incorporated social values too late. -4 -2 25. Need to take longer view towards monitoring. -1 0 26. OAs prescriptive uses does not work well. -1 -3 27. Needs to be implementable . . . affordable. 2 1 28. OAs are best management technique available. 2 -2 29. OAs balance polarized uses. 3 -1