The Effectiveness of Facilitated Discussions in REsolving Conflict at the New Bedford Harbor Suspended Site

This study examined a case of conflict between Federal decision-makers and local citizens regarding the selection of a clean-up plan for a coastal area contaminated by hazardous wastes. A prolonged dispute over remediation of the New Bedford Harbor Superfund Site resulted in a series of facilitated discussions intended to resolve the conflict. A descriptive survey was used to explore the effect of the discussions upon the risk perceptions of citizens and managers, thereby examining the effectiveness of the facilitated discussions in resolving conflict at the Site. The study found that the facilitated discussions did not cause risk perceptions to converge, . and concluded that the discussions averted, rather than resolved, conflict between citizens and managers at the New Bedford Harbor Superfund Site.


LIST OF FIGURES
. The cost of cancelling that one contract has not yet been negotiated, but will undoubtedly be considerable (MacNeil 1994). The Acushnet River is a small estuary in Southeastern Massachusetts, opening onto Buzzards Bay. The mouth of this estuary forms a well-protected harbor, New Bedford Harbor, which has been used for commercial purposes since English settlement in the early 18th Century. By the mid-1700s, the river had become an important center of shipbuilding and whaling, and the towns on its banks--New Bedford, Acushnet, and Fair Haven--prospered and grew. By the middle of the 19th Century, New Bedford had become the largest whaling port in the world (Ellis 1892).
Manufacturing also began in the 18th Century in the New Bedford area, albeit on a small scale. During the Nineteenth Century, however, as the whaling trade began to decline, the great capital which had been accumulated by whaling interests began to be shifted into manufacturing; while textile mills made up the most important segment of this industry, a great diversity of production facilities arose (Hegerty 1959) . In the late 19th and early 20th centuries, dozens of these vast mills were built in brick. With the mills came the tenements, or "triple-deckers:" three-story frame structures built to house the laborers who, with their families, migrated to New Bedford to fuel this industrial expansion. The mills--many vacant now, or partially so--and 3 the tenements dominate the areas surrounding the Acushnet River to this day.
Architecture is not the only legacy which New Bedford's industrial growth left upon the area's environment. From the 1940's until the late 1970's, manufacturers of electrical equipment dumped industrial wastes into the Acushnet River, New Bedford Harbor, and adjacent waters.
The wastes contained a number of toxic organic and inorganic substances, including: polychlorinated biphenyls (PCBs), polyaromatic hydrocarbons (PAHs), lead, cadmium, chromium, and copper (Pruell et al. 1990;USEPA 1992cUSEPA , 1989. Resulting levels of contamination in estuarine sediments ranged as high as 200,000 parts per million (ppm) PCBs 1 2 These high-end estimates presuppose almost impossible circumstances.
In an era which has seen the flight of "smokestack industry" from many of its former strongholds, contaminated marine sediments are just one of an array of difficulties faced by the towns adjoining the Acushnet River. New Bedford, in particular, is among the poorest municipalities in Massachusetts; demographic statistics present a picture of a metropolitan area which is disadvantaged in numerous ways.
According to the 1990 Census, the population of New  (Hird 1993;Johnston and Nixon 1992) In 1986 at a million dollars or more (Lewis 1990).

Lower Harbor
North Dartmouth

Upper Buzzard's Bay
The New Bedford Harbor Superfund Site.
Source USEPA 1989a. 13 The Hot Spot comprises the first OU, while less contaminated areas throughout the estuary, harbor, and upper bay comprise the second (USEPA 1989a;1992a;1992c)  Source: USEPA 1992a . 16 The decanted solids would then be incinerated in a hazardous waste incinerator 5 temporararily assembled on the west bank of the Acushnet. The residual ash would be temporarily stored in the CDFs, solidified if necessary. The incinerator and CDFs would be located at the foot of Sawyer street, in the North End of New Bedford, an area, like much of the city, of mills and tenements, largely residential.
According to the Hot Spot Proposed Plan, once the Hot spot sediments were dredged and incinerated, USEPA would move on to remediate the less highly contaminated areas of the Superfund Site (USEPA 1989a) . The Hot Spot Proposed Plan was ratified by means of a Record of Decision (ROD) which USEPA published in April, 1990, formally selecting the dredging and incineration as the means of remediating the contamination of the upper Acushnet River (USEPA 1989b) In 1992, USEPA issued an "Explanation of Significant Differences" (ESD) which modified the ROD, stating that the incinerator ash would be permanently, rather than temporarily, stored in the shoreside CDFs at the foot of Sawyer Street (USEPA 1992b). Initially, the Hot Spot remediation process was expected to take one year and to cost approximately $14.3 million (USEPA 1989a); this figure was revised to $16.1 million several years later when the plan was modified by the ESD (USEPA 1992b;1992c) .
In 1983 and 1984, the US Department of Justice filed civil complaints against five of New Bedford's electrical parts manufacturers for recovery of clean-up costs and natural resource damages (Casner et al. 1990). The suits were settled by a series of consent decrees from 1990 to l992, in which the five PRPs agreed to pay a total of $110 million to USEPA. Of this amount, $84 million would go toward the cleanup; $20 million for restoration under CERCLA's natural resource damages provision (42 U.S.C. section 9607(f) (1)); and the difference to reimburse the state and Federal Government for costs incurred (Allen 1992a;Garmon 1993 broadly empower the Agency to recover costs from polluters for cleanup of such sites (Williams 1993;Johnson 1990) Within this framework, the Act seems to have an uneasy regard for public participation in Superfund decisionmaking.
It requires a degree of public involvement in USEPA's selection and implementation of CERCLA response 6 plans, but is less than unequivocal regarding the relative importance of citizen input in the formulation of plans.
CERCLA contains provisions for citizen suits, but such suits are subject to important jurisdictional limits. The courts have frequently been called upon to determine the specific nature of these limits ~-and therefore the degree to which citizens have the right to challenge CERCLA response decisions.
In most cases, Federal courts have held that CERCLA deprives them of jurisdiction to hear any challenge, 6while the language of CERCLA distinguishes between "removal" and "remediation" actions, such distinction is not germane to this study; therefore, I will use "response" to encompass both removal and remediation of hazardous wastes Under CERCLA. by any party, to cleanup activities at a Superfund site until such activities have been completed.

I. D. 1. Statutory requirements.
As amended, CERCLA requires USEPA to engage in a range of public involvement activity in the process of cleaning up a superfund site. The Agency is required to solicit public comments during the cleanup and to keep the public fully informed throughout the process. The Act authorizes the Agency to provide citizen's groups with Technical Assistance Grants (TAGs) of up to $50,000 . "to obtain technical assistance in interpreting information" regarding almost any aspect of a Superfund Site and its cleanup. Among other things, the TAGs are intended to "facilitate public participation in the selection of [a] remedial action at the facility" and to "facilitate public participation at all stages of a remedial action" (42 U.S.C. Sxn. 9617).
The law requires USEPA to provide the public with "a reasonable opportunity to comment and provide information about [a] plan" and "an opportunity for a public meeting in the affected area." It requires that USEPA respond to public As has increasingly happened in recent years when USEPA ha~ chosen to employ incineration to remediate contaminated soils. Arguably, CERCLA/SARA as written failed to foresee the subsequent proliferation of citizen suits seeking to block USEPA remediation plans.

Intervention in Superfund consent decrees.
The Comprehensive Environmental Response, Compensation, and Liability Act of 1980 provides for intervention as a matter of right by "any person" who "claims an interest" in a superfund action (42 U.S.C. 9613(i) Of the eight, only Acus~n~t IV and VI bear directly or indirectly upon public Phart7c1pation; therefore, only these two will be discussed ere1n.  432 U.S. 33, 97 S.Ct. 2434, 53 L.Ed.2d 383 (1977. it1ng, accord, Conservation Law Foundation of N.E. v. e1lly, 950 F.2d 38 (1st Cir. 1991).

29
The jurisdictional bar of CERCLA section 9613(h) was written into the law with the 1986 SARA reauthorization .
According to Williams (1993), it appears to be a  27 While the focus of this study is on public participation an~ citizen suits, and while the Sxn. 9613(h) (4) exception which most concerns us relates to citizen suits, there a good deal of precedential overlap between CERCLA challenges brought by.citizens, PRPs, and state governments.
Therefore, while thi~ discussion will center upon citizen suits, relevant actions brought by other types of parties must be mentioned, as well.
citing the US Supreme Court in Block v. Community Nµtrition Institute, 28 the Court noted that the extent to which a statute precluded judicial review was determined not only by the language of the law, but by its structure and nature, as well as its legislative history. Delving into CERCLA's history as well as the law's intent, the District court concluded that both the legislative history and the language of CERCLA suggest that to allow judicial review of a ROD by an entity which may (or may not) be the subject of a subsequent recovery action would frustrate Congress' intent to provide a mechanism whereby hazardous sites can be neutralized expeditiously.
Moreover, the Court found, some of the events of this particular case were "suggestive of the bog into which courts would descend if CERCLA were interpreted to permit [PRPs] to obtain judicial review of the issuance of a ROD." The Court also considered the limited precedent which had developed following the passage of the 1982 law, citing opinions which had found that courts lacked jurisdiction to review PRP claims brought to enjoin Superfund cleanup activities. 29 Supp. 405 (N.D.Ill. 1984) and Aminoil v. United States Environmental Protection Agency, 599 F. Supp. 69 (D.Calif. 1984).
~584 F. Supp. 1005(N.D.Ohio, 1984, affirmed 767 F. 2d 263 (6th Cir. 1985). the right to obtain judicial review of a Superfund ROD. The ~eter... §. court had ruled the issuance of a ROD by USEPA a "final administrative action" subject to judicial review under CERCLA; moreover it found the PRPs' potential liability for the clean-up to comprise the "actual or threatened injury" required to establish standing under Article III. 31 The reasoning which the Lone Pine court relied upon to reject the precedent of Peters has become a fundamental justification for denying PRPs review of USEPA RODs: that the PRPs will have the opportunity to state their case in the subsequent CERCLA cost recovery suit, and that therefore the ROD is not in and of itself "a final agency action." In the words of the Court: [There is] no reason why plaintiffs cannot raise as a defense in a cost recovery action every objection to the ROD which they could legitimately raise in a judicial proceeding at this time.
The plaintiffs' motion was accordingly dismissed for lack of jurisdiction.
It is important to note that the reasoning which the timately, however, the Peters court dismissed PRPs' suit for failure to state an actionable claim.

32
The the Pine court acknowledged as much when it stated, dictum, k_l.,!.!o~n~eS:.-:..==--that the statute may contemplate a different rule of judicial review in the case of a victim of a hazardous waste site. The statute is designed particularly to protect such persons and, unlike the persons responsible for the hazardous waste, no specific provision is contained in the statute under which they can obtain judicial review ... In any event, it is unnecessary to resolve that question now.
In codifying the decision of the Lone Pine court by means of CERCLA During the remedial investigation and feasibility study, [USEPA] shall solicit the views and preferences of the affected community on the remediation and disposition of hazardous substances, pollutants, or contaminants at the facility.
[These] views and preferences ... shall be ... considered in the screening of remedial alternatives for the facility.
This bi-directional information flow was emphasized within the revision's public meeting requirements, as well: Public meetings shall be designed to obtain information from the community and disseminate information to the community concerning [USEPA's] facility activities and pending decisions. (US House of Representatives, 1994) . 48 While the Senate bill, S. 1834, did not elaborate upon public participation in Superfund to the extent that the House bill did, its effect would have been substantially the same. Like the House version, it would have required USEPA to employ frequent public meetings and would have established Citizen Information and Access Offices. Also like the House version, and in contrast to current law, S .   (Collins 1993;Shaw 1994) . Frank also lobbied the White House, involving the General Accounting Office as well as Leon Panetta, President Clinton's budget director (Glass and Corey 1993).
The way had been cleared, legally, for USEPA Region I to continue with its incineration plans. However, the regional office was strongly encouraged behind the scenes, by the Administrator as well as by Judge Young, to attempt to settle the dispute through mediation (Garmon 1994;Shaw 199 4 Much recent research has suggested that locational disputes over hazardous facilities result, in part, from differences in risk perception between the "risk professionals" who make siting decisions, and the public who must live in the shadow of such facilities (e.g., Cvetkovich and Earle, 1992;Slovic, 1987). Uncertainty is inherent to the science of risk assessment (Patton, 1993). To complicate matters in New Bedford's case, there is no "zerorisk" option; the choice lies between the risks presumably inherent in the existing contamination in the waterway, versus a set of equally uncertain potential risks resulting from various cleanup options, such as incineration.
Alternative Dispute Resolution (ADR) is an umbrella term which encompasses a variety of structured discussion techniques which have been used increasingly in recent years to attempt to resolve environmental disputes in a consensual, non-litigious manner (Fischer, 1993;Mosher, 1983) . In the New Bedford case, ADR has taken the form of "facilitated discussions," coordinated by the Massachusetts Office of Dispute Resolution (ODR) .
If, as the literature suggests, differences in risk perception between professionals and the public are a source of environmental disputes, then one definition of effectiveness in resolving such disputes would be the 57 reduction of such differences.
The problem which this study examined, therefore, was whether ADR had, in this instance, reduced differences in risk perception between "risk managers"--environmental professionals--and "stakeholders"-residents of the towns adjoining the Acushnet River.
In order to measure perceptions of a variety of risk and trust issues related to the contamination of the

II. B. 1. Risk perception, assessment, and uncertainty.
Human activities give rise to a broad and diverse range of technological hazards: from minor releases of pollutants to catastrophic nuclear and chemical plant accidents; from automobile accidents to the collapse of dams (Cutter 1993).
Quantitative risk assessment is fundamental to governmental environmental decisionmaking (Patton 1993;Russell and Gruber, 1987). In the case of carcinogenic and toxic chemicals, risk assessment uses extrapolations from animal data to attempt to estimate the average probability that a particular consequence will result from a given level of exposure to a particular substance. Probabilities so obtained are used by USEPA as the foundation for risk management, the process by which the Agency decides whether and how a risk should be regulated or mitigated (Patton 1993 ). While risk assessors work to err on the side of conservatism, the methodology of risk assessment necessarily employs a good deal of inference, extrapolation, and assumption, so that quantitative risk assessment is inherently uncertain (Patton 1993;Sheuplein 1993;Russell and Gruber 1987) . According to Freudenburg and Pastor (1992), "[q]uantitative estimates of overall risk are likely to involve the tyranny of illusory precision." Moreover, recent behavioral research has shown that expert risk assessments are built largely upon subjective judgements (Gardner and Gould, 1989). If one accepts these points of view, then, the truly objective risk quantification becomes an impossibility.
To the extent that federal decisionmaking sees risk as something which can be accurately quantified, it takes a technical, or objectivist, view of risk.
In contrast to this model is one proposed by researchers in the behavioral sciences, which might be referred to as the constructionist, or subjective, view. According to this view, risks and hazards are broad social phenomena, functions of collective actions and judgements with respect to the physical world.
Within this paradigm, perception becomes more important than efforts at quantification (Cutter, 1993;Cvetkovich and Earle, 1992), while "risk" refers not just to chance of death or loss of "statistical lives," but incorporates an array of undesirable effects (Gardner and Gould, 1989) .
The study of risk perception originally concerned itself with natural hazards, but has since evolved to deal largely with technical ones. The work may be divided into two broad categories: that which utilizes a primarily psychological approach, and that which is essentially spatial. The former is characterized by emphasis on cognitive aspects of risk perception (often utilizing controlled experiments) while the latter is more functionally oriented, generally site-specific, and often concerned with the relationship between behavior and political or physical action (Cutter 1993). Starr (1969) was among the first to attempt to quantify public perception of technological risks, employing a "revealed preference" methodology to draw conclusions about the risks and benefits of technologies based upon historical data (Slavic 1987) . This approach was superceded by what Slovic refers to as the · "psychometric paradigm:" the use of "expressed preferences" or survey methodology to show that different groups perceive given risks in quantifiably different ways, and that, conversely, different risks can be distinguished by quantifiable characteristics using factor analysis (Slovic 1987) .
In the groundbreaking Risk and Culture (1982), however, Douglas and Wildavsky offer a far more subjective view of risks and risk perception. Drawing on anthropology as well as political theory, these authors argue that risks are "cultural constructs." That is, while risks are often real (though frequently hidden) , their perception and selection are functions of particular cultural biases.
By 1990, Wildavsky and Dake could describe three major theories of risk perception prevalent in the social sciences. "Knowledge theory" posits the idea that people perceive risks because they actually know things to be dangerous. "Personality theory," conversely, proposes that risk perception is a function of one's personality, while "economic theory" sees a personal cost-benefit calculus driving risk perception. These researchers conclude that personality variables tend to be the best predictor of attitudes toward environmental risks. Similarly, Samdahl and Robertson (1989) examined the existing body of work as well as their own questionnaire, concluding that while socioeconomic variables are generally poor predictors of environmental concern, political ideology corresponded strongly with pro-environmental values.
Other researchers have examined the psychological context of complex environmental decisionmaking, including that concerned with technological risk. A number of researchers have shown that, faced with a great deal of complicated information, individuals--experts included-employ "cognitive oversimplification" or heuristic methods to arrive at a decision (Miller 1985, Varis 1989. Miller (l985) shows that such simplification inevitably results in decisionmaking bias, distinguishing between "cognitive bias"--that which results from the simple inability to incorporate all relevant factors into a decision--and "motivational bias"--that which is driven by ulterior motives.
In contrast to these psychological approaches to the study of risk perception, geographers have taken a more functional or contextually-oriented approach, examining the factors influencing individual or group response to actual situations (Cutter, 1993). Furuseth (1990) Much recent work has shown that differences in risk perception between professionals and laypeople are a primary source of conflicts over environmental technology (e.g. Fort et al. 1993;Cvetkovich and Earle 1992;Gardner and Gould 1989;Slovic 1987;Miller 1984). Miller (1984) has shown that this disparity results, in part, from inherent biases held by experts and laypeople, including biases resulting from media coverage and attitudes of distrust toward certain categories of individuals, such as representatives of government. Moreover, the two groups may define "risk" in this context in entirely different ways (Slovic, 1987).
Professional environmental managers tend to define risk in terms of probability of death or injury while laypeople employ a broader calculus, taking into account more qualitative factors. These include the existence of disagreement among experts; the degree of equity with which risks are borne (Gardner and Gould 1989); effects upon property values (Cvetkovich and Earle 1992); and the degree of public control over the selection of risks (Fiorino 1989 The two terms regard the same phenomenom from slightly different perspectives. Those who use NIMBY "tend to see only the location, not the technology, as problematic," while users of the term LULU implicitly acknowledge that certain types of facilities are likely to meet with public opposition regardless of their proposed location (Freudenburg and Pastor 1992). Hazardous waste treatment facilities are among the least . wanted (Popper 1981 which is genuinely comparative --treating as problematic not just the views of the general public, but of both the public and the technical community (1992) .

II. B. 3. Public participation in environmental decisionmaking.
A subtext of much recent risk perception literature-particularly that written from the social science perspective--is the desirability of increasing local public involvement in environmental decisionmaking.
Goals of increased participation include enhancing the acceptability of LULUs as well as the furthering of democratic ideals (Kasperson et al. 1992;Fiorino 1989;Kasperson 1986;Portney 1985) . Arnstein (1969) pointed out that there is a hierarchy of citizen involvement in government decisionmaking, ranging from the nominal ("tokenism") to the actual ("citizen power"). An objective of some risk perception studies has been to forge . a theoretical link between perception and participation, i.e., to explain citizen action (Cutter 1993).

II. B. 4. Alternative dispute resolution.
Clearly, a priori public involvement is more efficient than a protracted environmental dispute (which is not to suggest that either is inevitable, nor that the two are mutually exclusive). Once a dispute does occur, it may be resolved either by legal proceedings or by alternative 66 dispute resolution (ADR) . Cormick (1980) Hird (1993) agreed that Superfund sites are more likely to occur in communities with more non-white residents, but found a positive correlation between wealthier communities and the 67 frequency of hazardous waste sites.
Hird's research found great variability in the cost-effectiveness of Superfund, concluding that costs per statistical life saved ranged from $ 34 0,000 to $77 billion. The limit of cost-effectiveness for environmental clean-ups is generally considered to be $l0 million per statistical life saved (Hird, 1993 1. Those who opposed incineration: local citizens and their representatives--the three ad hoc environmental groups as well as local politicians .

2.
Those who favored incineration: Federal and State coastal resource managers--specifically, representatives of USEPA and DEP .

III. A. 2. Study groups·.
In order to test the general hypothesis at a discrete point in time, the two disputant groups were subdivided so that differences in risk perception between citizens and managers within ODR's discussions might be compared with differences in risk perception between citizens and managers uninvolved in the discussions. In other words, the following four sub-groups of stakeholders were regarded for analysis: 70 up 1: Citizens' representatives who participated in the Gro facilitated discussions chaired by ODR.

Group 2:
Citizens of the Greater New Bedford Metropolitan Area who had not participated in these facilitated discussions, represented by a random sample.

Group 3:
Coastal resource managers included in the facilitated discussions chaired by ODR.

Group 4:
Coastal resource managers who were uninvolved in these facilitated discussions, represented by a nonrandom sample .

III. A. 3. Assumptions.
In order to establish the context of the hypothesis, the research employed the following assumptions:

Assumption 1:
It is assumed that the current perceptions of Group 2, regarding hazards associated with the New Bedford Harbor Superfund Site and its remediation, are statistically similar to the perceptions of Group 1 at the point at which these citizen representatives became involved in ODR's facilitated discussions.

Assumption 2:
It is assumed that the current perceptions of Group 4, regarding hazards associated with the New Bedford Harbor Superfund Site and its remediation, are statistically similar to the perceptions of Group 3 at the point at which these coastal resource managers became involved in ODR's facilitated discussions.

III. A. 4. Major Hypothesis.
It was hypothesized that the involvement of individuals in the series of facilitated discussions chaired by ODR reduced the differences in hazards perception between the two opposing sides in the dispute over the remediation of the the New Bedford Harbor Superfund Site. 1) The difference in perception between Groups 1 and 3 is smaller than the difference in direction between Groups 2 and 4.

2)
The two public groups (Groups 1 and 2) lie at one end of the continuum, while the professional groups (3 and 4) lie at the other.
In other words, while the facilitated discussion has brought the two sides closer together, it has not changed the It is hypothesized that there are no statistically significant differences in the measured perceptions of the four groups, regarding the hazards associated with the New Bedford Harbor Superfund Site.
It is hypothesized that there are statistically significant differences in the measured perceptions of the four groups, regarding the hazards associated with the New Bedford Harbor Superfund Site.
The research hypothesis is in turn subdivided into three one-tailed sub-hypotheses, as follows:

sub-hypothesis 1:
It is hypothesized that the difference in perception between Group 1 and Group 3, regarding the hazards associated with the New Bedford Harbor Superfund Site and its remediation, is, to a statistically significant degree, less than the difference in perceptions between Group 2 and Group 4.

Sub-hypothesis 2:
It is hypothesized that the difference in perception between Group 1 and Group 2, regarding the hazards associated with the New Bedford Harbor Superfund Site and its remediation, is, to a statistically significant degree, less than the difference in perceptions between Group 1 and Group 4.

Sub-hypothesis 3:
It is hypothesized ·that the difference in perception between Group 3 and Group 4, regarding the hazards associated with the New Bedford Harbor Superfund Site and its remediation, is, to a statistically significant degree, less than the differences in perception between Group 3 and Group 2.
The three sub-hypotheses may be graphically represented in relation to the main hypothesis as follows: It was immediately apparent that there was a great deal of interest in the New Bedford Harbor Superfund Site among local people; it seemed, therefore, that residents did not regard the ten-page survey to be as onerous as they otherwise might have. Generally, the pretest respondents seemed to find that the survey required some thought, but was not difficult, although they found some questions to be in need of clarification. These were revised in the subsequent, and final, version.
During the spring and summer of 1994, the New England March of 1994; many expressed a belief that the Government was trying to put them out of business (Pollack 1994a and1994b

III. D. 1. Discussion participants: Groups 1 and 3.
Groups 1 and 3--citizens and professionals involved in ODR's facilitated discussions--were queried comprehensively. The Office of Dispute Resolution provided a list of forum participants and affiliations; based on group affiliations, participants were divided into citizen representatives versus professionals for study purposes.
Facilitators were excluded from the study group, while elected officials were placed in Group 1.
The discussion facilitators introduced me at one of the last of the forum's regular meetings, in July, 1994. I contacted each of the discussion participants soon thereafter, briefly describing the nature of the study, and asking them to complete a survey, with an assurance of confidentiality. I emphasized that I was looking for individual perceptions rather than official position statements. In all cas~s, the forum participants agreed to at least look at the survey, although some of the professionals expressed a concern that their responses would be kept truly confidential (one went so far as to say "I could lose my job over this"), while others communicated an initial unwillingness to separate their own ideas from that of their employer. Subjects were then mailed surveys (public or professional, as appropriate) along with a personalized cover letter and return envelope.
Survey response rates for Groups 1 and 3 were as follows: The list of participants in the facilitated discussions when the surveys were administrated may be found in Appendix C.

III. D. 2. Professional non-participants: Group 4.
The study's Group 4 (marine resource managers uninvolved in ODR's facilitated discussions) was chosen in a non-random, non-comprehensive manner. Admittedly, this is a weak selection method in terms of· validity; however, it was the only practical way to assemble an adequate number of individuals with the required characteristics. Subjects for this segment of the study were choseri from my own acquaintances in the field of environmental management and these professionals' referrals. Criteria for selection were as follows: 1.
That the individual was involved in any aspect of environmental or coastal management on a professional level.
2. That s/he was working in Massachusetts, and professed at least a general knowledge of the New Bedford Harbor Superfund Site.
3. That s/he had at no time been an active participant in ODR's facilitated discussions . 83 In order to ensure that Group 4 was approximately the same size as Groups 1 and 3, nine individuals were contacted.
Methodology from this point was essentially the same as for Groups 1 and 3: after outlining the purpose of the study, I asked whether they would be willing to participate.
In most cases, they initially agreed. A copy of the professional version of the survey was then mailed to each, along with a personalized cover letter and return envelope. After reading the survey, one member of Group 4 declined to participate: an employee of NOAA, who simply said that he did not feel that his completion of the survey would be "appropriate." Distribution and returns were as follows:

III. D. 3. Public random sample: Group 2.
In order to query the non-aligned New Bedford area public (Group 2) I the public version of the survey was administered to a stratified random sample of residents of New Bedford, Fair Haven, and Acushnet from 6 -27 July, 1994. Using the approach described by West (1993), ten random traverses were laid across a map of the study area (Champion Map Corporation, 1990) . City blocks were selected at random intervals along these lines; rural areas were 84 excluded. The distribution and return of surveys to , and from, Group 2 is fully described by Table 3 . 3.

IV· A. Description of sample by group.
As explained in Chapter III, the study's survey was distributed to four groups: Group 1: Citizens' representatives who had participated in the facilitated discussions chaired by the Massachusetts Off ice of Dispute Resolution (ODR) . citizen-participants were queried comprehensively.

Group 2:
Citizens of the Greater New Bedford Metropolitan Area who had not participated in these facilitated discussions, selected by means of a stratified random sample of New Bedford, Fair Haven and Acushnet residents.
Group 3: Coastal resource managers who had been involved in the facilitated discussions. Manager-participants were, like Group 1, queried comprehensively.

Group 4:
Coastal resource managers who had not been involved in these facilitated discussions, represented by a non-random sample.
Distribution and return of surveys by group are summarized in  Those who further described their employment were about evenly divided between blue and white-collar workers.
Again, no single source of information "concerning local events and controversies" seemed most important: radio, television, and newspapers were all rated highly by this group.
Predictably, in comparison with Group 1, far fewer Group 2 respondents considered public meetings to be an important source of information: 14.3% as opposed to 87.5% among Group 1.
Group 3: Not a single Group 3 . respondent was a resident of the New Bedford area. Reported period of involvement in environmental management ranged from 2 to 26 years, with a mean of approximately 13.4 years. When these individuals were asked how often they interacted with non-professionals during the course of their work, responses showed no clustering, ranging from "at least once a day" to a minimum of "once a month" or "varies too much to say." Asked to describe their role in site management, over half of these professionals (62.5%) considered their primary role a managerial one; l~ss than a third ( Group 3 (mean response 7 . 71); they were slightly less keen than Group 3, however, on the desirability of expanding public involvement in Superfund (mean response 6.83).

IV· · ·
Results of each of the survey's thirty-three "core questions" {Questions 8-40, those which were administered to all four study groups) were examined in order to ascertain whether the hypothesized relationship was present. The reader will remember that the study's hypothetical model was one where:

1.
The difference in perception between those groups which had participated in the facilitated discussions (Groups 1 and 3) was smaller than the difference between those which had not (Groups 2 and 4), but 2. Direction of differences separating citizens from managers remained constant; that is, the perceptions of the citizens (Groups 1 and 2) occupied one end of a continuum, while those of the managers (Groups 3 and 4) occupied the other.
This hypothetical relationship is diagrammed, once again, below.
Where the expected relationship was generally evident, statistical testing was undertaken in order to determine whether differences among groups were statistically significant.
In other words, if the graphic relationship expressed in the study's hypothetical model was initially discernable, and there were significant differences among the study groups, then the research hypothesis was accepted, and the null rejected. Where the graphic relationship of the hypothetical model was not evident, then, the null hypothesis could be accepted, and the research hypothesis rejected; questions which fell into this category were not initially examined for statistical significance.

IV. B. 2. Statistical approach: Likert-scale questions.
In keeping with the methodology explained above, the first step involved a simple examination of the means of the Likert score by group for each question . Where the hypothetical relationship was initially apparent, the Student-t test was used to test for significant differences among group means.
Since Group 2's mean dominated the study's universal mean by virtue of Group 2's proportionally large n (n 2 = 110 while N = 134), the three smaller groups were tested for divergence from the universal mean, µ. Strictly speaking, the existence of statistically significant differences between each of the smaller groups and µ would not imply d ' f i ferences among the smaller groups (e . g., between 3 and 93 4) .
However, in the context of the limitations imposed by the differences in size among the study groups, the relationship between µ and the three smaller groups (Groups 1 , 3 , and 4) was taken to be an indicator that the hypothetical relationship was present in some form.
conversely, the lack of such significant differences allowed the research hypothesis to be rejected, and the null accepted.
The formula for the Student t Test is as follows: Where:  Blalock's (1979) table for the distribution of t in a twotailed test was used to determine whether or not differences among groups were statistically significant at the 95th percentile, i.e. a s .OS.

IV. B. 3. Statistical approach: Frequency-based responses.
The first step of hypothesis testing of the survey's checklist and rank-order responses, like that for the Likert-scale questions, involved simply examining the margin totals for each checklist or rank-order sub-question in order to determine whether the general hypothetical 94 relationship was apparent. Where it was, statistical analysis of both these types of questions focused upon categorical frequencies.
In the case of checklist questions, responses for each group were divided into two sub-groups based upon whether the respondent had checked a particular line (considered a "yes" response) or not (considered a "no"); non-respondents were omitted from analysis. For rank-order questions, the number of respondents who had assigned a specific rank to a particular sub-question was measured against those who had not.
Following Siegel and Castellan (1988), the Chi-Square (X 2 ) test, corrected for continuity with the data cast into a 2x2 contingency table, was used to test for differences among groups. Again, the considerably larger n of Group 2 required an asymmetrical approach to the testing. Response frequencies for each of the smaller groups were tested against those of the largest group (Group 2). Again, the existence of significant differences between each of the three smaller groups, an~ the larger group, was used as a general indicator, while the lack of such differences was considered to be evidence for acceptance of the null hypothesis.
The mean was disregarded for this set of analyses.
The contingency table used for analysis is presented in  Contingency table for X 2 analysis of response frequencies.
Group 1, 3, or 4 Group 2 Frequency "no" Cell "A" Cell "B" Frequency "yes" Cell "C" Cell "D" The formula for the X 2 test corrected for continuity is as follows:  o, "No current health risk," to 10, "Great current health risk." A score of 5 was marked "Moderate health risk." Results and analysis are summarized in Table 4.3.  Examination of survey response margin totals exposed a relationship which appeared to be far more prevalent than the hypothesized one. This relationship may be described as It is hypothesized that there are no statistically significant differences in the measured perceptions of the three groups, regarding the hazards associated with the New Bedford Harbor Superfund Site.

~:
It is hypothesized that there are statistically significant differences in the measured perceptions of the three groups, regarding the hazards associated with the New Bedford Harbor Superfund Site.
The sub-hypotheses used to test for polarization were as follows:

Sub-hYPothesis 1:
It is hypothesized that the perceptions of Group 1 and Group 3, regarding the hazards associated with the New Bedford Harbor Superfund Site and its remediation, lie at either extreme of the perceptual continuum formed by the perceptions of all four groups.

sub-hypothesis 2:
It is hypothesized that the perceptions of Group 1, regarding the hazards associated with the New Bedford Harbor Superfund Site and its remediation, are different, to a statistically significant degree , from those of Group 2 (for frequency-based questions), or from the universal mean, µ (for Likert-scale questions) .

Sub-hypothesis 3:
It is hypothesized that the perceptions of Group 3, regarding the hazards associated with the New Bedford Harbor Superfund Site and its remediation, are different, to a statistically significant degree, from those of Group 2 (for frequency-based questions), or from the universal mean, µ (for Likert-scale questions) .
It will be noted that, again, µ and Group 2 were treated as, if not the same, similar for purposes of analysis, since the proportionally large n of Group 2 caused it to dominate the study's statistical universe .
The method for testing the polarization hypothesis was similar to that used to test the major hypothesis . Again, in order to validate the model, the research first examined margin totals for the existence of the general polarized relationship. Where thi~ relationship was graphically apparent, statistical testing was undertaken to test for significant differences. Acceptance of the research hypothesis required SSDs between Group 1 and the mean (or its proxy, Group 2), as well as between Group 3 and the mean (or, again, Group 2). Though Group 4 was tested along with the other groups, its statistical status was in fact irrelevant to the search for polarization. of the three types of questions (Likert, checklist, and rank-order) used in the survey's "core" section, approximately 33% of Likert-scale questions (7 of 21) , 24% of checklist sub-questions (9 of 38), and none of the rankorder sub-questions showed the general polarization relationship. Again, the Student-t Test was used to analyze the results of the Likert-scale questions; the three smaller groups were each tested for significant differences against the universal mean, µ, dominated by Group 2.
If Group 1 and Group 3 each showed statistically significant differences from µ, then the model was considered validated, and Hr was accepted.
Once again, frequency-based questions (rank-order and checklist) were tested by means of the Chi-square test corrected for continuity. Each of the three smaller groups was compared, statistically, with Group 2; where Group 1 and 3 were both found to be statistically different from Group 2, ~ was, again, accepted. O on the Likert scale was marked "They know very little;" 5 was marked "They know something;" 10 was marked "They know a great deal." Results and analysis are presented in Table 4.5.  Question 17: This question asked respondents to rate the risks of incinerating the contaminated harbor sediments, using a Likert scale whereon 0 was marked "High risk;" 5 was marked "Moderate risk;" and 10 represented "No risk." Once again, a polarized relationship between Groups 1 and 3 was graphically apparent. Since, however; the standard deviation for Groups 1 and 2 were larger than the means for those groups, the validity of statistical analysis using the parametric t-Test was called into question. Therefore, results alone follow in Table 4.7.  scale was marked "They know very little;" 5 was marked "They know something;" 10 was marked "They know a great deal."

IV. D. 1. Analysis of Likert-scale responses: Polarization.
Results and analysis are presented in Table 4.8.

Question 40:
Question 40 asked respondents to compare the health risks stemming from the harbor contamination with those from the proposed incineration, using a Likert scale whereon O was marked "Incineration is greatest risk;" 5 was marked "Equal risks;" and 10 was marked "Existing contamination is greatest risk." Standard deviations andµ were too large to permit use of the t-Test; results alone are therefore presented in Table 4.11.
An examination of margin totals suggested that the results of a number of the survey's checklist questions might fit the polarization model. Where the polarized relationship was initially apparent, statistical analysis was undertaken using, once again, the X 2 test corrected for continuity.
Question 9: As mentioned earlier, Question 9 asked respondents to identify whom they considered to be "at risk" from the contamination at the New Bedford Harbor Superfund site by checking from a list of fourteen demographic and geographic categories. The polarization relationship appeared in the responses to three of these sub-questions: Question 9b: Question 9b specified "Old people" as a possible category of individuals "at risk" from the harbor contamination. A summary of results and analysis, using the x2 test to compare Groups 1, 3 ,. and 4 with Group 2 I is presented in Table 4.12.  Question 9d: Question 9d asked respondents whether they considered themselves to be "at risk" from the harbor contamination.
Results and analysis are presented in Table   4.13. -----since the response of only one of the three smaller groups showed a statistically significant difference from that of the large random sample (Group 2), the null hypothesis was once again accepted: it was concluded that the polarization model did not apply to the results of Question 9d . However, Group 3 was, once again, demonstrated to be a statistical outlier .
Question 9m: Question 9m asked respondents whether they considered "All residents of Dartmouth" to be "at risk" from the sediment contamination. Results and analysis are presented in Table 4.14.
The lack of statistically significant differences among group responses to Question 9m required that the null hypothesis be accepted in this instance.
Question 10: Question 10 asked respondents to choose from a list of nine ailments in order to express what they considered to be "the health risks of the current contamination." The responses to two of these nine subquestions appeared to fit the polarization model; thus statistical testing, summarized in the tables below, was undertaken.

Question lOa:
Question lOa presented "Cancer" as a possible risk of the harbor contamination . Results and analysis appear in Table 4 . 15.
The null hypothesis was accepted in this instance, as a result of the lack of statistically significant differences among group responses.
Question lOc: Question lOc listed "Respiratory problems" as a possible risk of the harbor contamination. Results and analysis are presented in Table 4.16.
Again, the research hypothesis was rejected, and the null accepted, due to the lack of significant differences among group responses to Question lOc. Group 3 was once again the sole statistical outlier.
Question llf: Question 11 utilized a checklist naming six possible pathways to ask respondents how health risks might be transferred from the Superfund site to the population .
Of the six sub-questions, only one appeared to support the polarized relationship: sub-question llf, which referred to "Eating garden vegetables." Results and analysis appear in Table 4.17.
The null hypothesis was once again accepted for the results of Question llf.
Question 26: This question offered respondents a checklist of eight possible outcomes of the government's clean-up of New Bedford Harbor, asking them which results they would prefer . Three of the sub-questions which appeared to show the polarized relationship were tested for significance; analysis and results are summarized in the following tables.

Question 26a:
This sub-question asked whether respondents would like to see the 11 Clean-up ... provide jobs. " The null hypothesis was once again accepted; again, Group 3 appeared as a statistical outlier.

Question 26c:
This sub-question put forth "restore commercial options of harbor" as a clean -up result.
Due to the inability of the analysis to show statistical differences between the study groups and the universal mean, the research hypothesis was rejected, and the null accepted, for Question 26c .
Question 26e: This sub-question asked whether respondents would like to see the clean-up "enhance tourism." ------Again, the research hypothesis was rejected, and the null accepted, for Question 26e, since none of the differences between groups proved significant .
As a result of the foregoing analyses, it was concluded that the polarization model was not present to a statistically significant degree in any of the survey's checklist questions. Nor was polarization discernable in the responses to any of the survey's rank-order questions .

IV. D. 3. Summary of analysis: Polarization.
Though initially apparent in 33% · of the study's core Question 13 was discussed above; results and analysis were presented in Table 4.5; no SSDs were demonstrable.
Question 20: Survey participants were asked how much they thought USEPA's scientists ''really [knew]" about the health risks of the proposed incineration. Question 20 was discussed above, and a summary of response and analysis presented (Table 4.8). Only Group 3 was statistically different from the universal mean.
Question 21: Participants were asked "to what extent" they thought USEPA "[told] the public the truth" about the risks of incineration. The question utilized a Likert scale, where 0 was labelled "They tell very little of the truth;" 5 was marked "They tell some of the truth," and 10 represented "They tell the truth 100%." Results and analysis are summarized in Table 4.21. ----In this instance, analysis showed that the professional groups each differed, statistically, from the public random sample, while the citizen-participant group (Group 1) did not.

IV. E. 2. Expectation of benefits from clean-up.
Question 22: An attempt was made to gauge respondents' expectation of benefits resulting from the Superfund cleanup of the Acushnet by asking "how clean, and how safe" respondents expected the river to be when the clean-up was 119 complete. The question used a Likert scale on which 0 was marked "As dirty and unsafe as before;" 3.5 was marked "Safe for swimming;" 7.5 was marked "Fish will be edible;" and 10 was marked "Shellfish will be edible." Response and analysis are summarized in Table 4.23. In this instance, no statistical differences could be discerned among groups.
Question 29: A different approach to measuring respondents' expectations of Superfund clean-up benefits was used in Questions 29 and 30, each of which asked a willingness-topay question regarding the harbor clean-up. Question 29 asked respondents how much they would "personally" be willing to pay to be sure that the Acushnet would be restored to a state where "fish and shellfish were safe to eat, and the river was safe for children to swim in." A Likert scale divided into ten increments was provided; these were marked with dollar values ranging from O to "Over $1000." Standard deviations were generally too large to permit the use of the Student-t test; therefore, results alone are presented in Table 4.24.   posited a scenario where the river was "no longer highly contaminated, but still not safe enough to swim or fish. 11 121 Again, due to large standard deviations, results are expressed in Table 4.25 without test results. ----Again, large differences between group means were apparent, in spite of the inappropriateness of statistical analysis to this set of responses.
In this instance, however, the public random sample (Group 2) least valued the posited outcome.
Within each group, there seemed to be a large difference between the value placed on the Question 29 scenario, and that associated with Question 30's hypothetical situation, though here, too, statistical confirmation was not possible. For Group 1, a cleaner river was worth 57% of what the pristine river was worth; for Group 2, the figure was less than 15%; and for both Group 3 and Group 4, the proportion was approximately 43%.

IV. E. 3. Knowledge of source of funds.
The last survey question for which responses were examined attempted to discern to what extent respondents were aware of the sources of funding of the Superfund clean-122 UP· As mentioned in an earlier section, corporations responsible for the contamination of New Bedford Harbor (potentially responsible parties: PRPs) have agreed to contribute $110 million toward the clean-up, although superfund itself provided federal monies for the initial investigations and planning, cost-recovery actions, etc.

Question 28:
Respondents were asked, simply, "Who will pay for this [Superfund] cleanup?" Written responses were divided into two sub-groups: those which recognized the contribution of the PRPs, versus those which expressed a belief that "government" or "the taxpayer" would bear the entire burden. Statistical testing for differences among groups was then undertaken in the same manner as for all other frequency-based responses, using the Chi-Square (X 2 ) test corrected for continuity and cast into a 2x2 contingency table.
Results and analysis are summarized in Table 4.26. It is notable that a large majority of the public random sample, Group 2, expressed a lack of knowledge of the most important source of funding for remediation of the New Bedford Harbor Superfund Site.

Conclusions.
A MaJ'or hypothesis.
The research has not supported the major hypothesis.
The study has concluded, therefore, that the facilitated discussions held by the Massachusetts Office of Dispute

V. B. 3. Trust in government.
The study found that the public groups shared a relatively low degree of trust in the ability of the government to determine and deal with the problems arising from the contamination and its remediation.

Responses to
Question 13 ( it. As rational actors under such circumstances, they will "take the devil they know," or think they know (the contaminated harbor) over the one that they do not (incineration) .

v. B. 4. Expectation of benefits.
Three of the survey's questions (22, 29, and 30) examined respondents' expectation of benefits from the harbor clean-up . While analysis of Question 22 (Table 4.23) found the response of Group 3 to be significantly higher than that of the other groups, no group felt that fish in the harbor, after clean-up, would be safe to eat; Group 2 expected the harbor to be just safe enough for swimming (mean Likert score = 3.6). Clearly, public expectation of water quality improvement as a result of the Superfund project was low . A. Citizens strongly oppose incineration; believe that emphasis of forum should be on finding alternatives.

Managers:
Incineration remains a viable option; cost of switch in technologies must be justified.

II
Economics of incineration and remediation I .

A.
Citizens believe hazardous waste incineration will cause economic harm to tourism and property values.
Managers believe that timely cleanup of Hot Spot is key to economic revitalization of the harbor.

B.
Citizens: Stigma of incineration will be permanent.
Managers: Incineration will produce short-term effect only.

A.
Citizens : Public health is more important than cost; managers seem too concerned with costs.
Managers: Public health is paramount but agencies are obligated to contain cost, as well.
B. Citizens: The safest and best alternative must be found, in spite of existing decisions and contracts. Managers express importance of relying on proven technology.

B· 1.
Coverletter for administration of survey to public random sample (Group 2), English.
(Researcher name, affiliation and address) New Bedford Harbor Superfund Site Survey Dear Madam or Sir: As you probably know, the sediments (mud) of the Acushnet River are contaminated with high levels of PCBs (polychlorinated biphenyls), particularly in the area north of the Coggeshall St. Bridge, between the towns of New Bedford and Acushnet. What to do about this contamination has been the source of great controversy over the past several years; the Federal Government, the State, and citizens' groups have all gotten involved in this debate. This is an opinion survey, from which we hope to learn how you, personally, feel about this contaminated area of the river, which the government calls the New Bedford Harbor Superfund Site. We hope that this survey won't take more than 15 minutes of your time.
Some of the questions may seem irrelevant, but please answer them anyway. There are no right or wrong answers; please read the questions carefully, then answer them in the way that best reflects your personal opinion.
The information which you provide will be kept COMPLETELY CONFIDENTIAL. Your name will not be given to any organization or company of any kind; you will not be put on any mailing lists. The surveys will be destroyed once the results are tabulated. Have you ever been active in any group or organization which was concerned with this Superfund site?    I----I----I----I----I----I----I----I----I----I----I  0  1  2  3  4  5  6  7  8  9  10 9. Who would you say is AT RISK from this contamination? (Please CHECK ALL that apply) . In your opinion, HOW might these risks be transferred from the river to humans?
(Please CHECK ALL that apply) Bre th' · a ing air ...... .   14. How did you, personally, learn about this contamination problem and its risks? (Please write in.) In 1990, the EPA decided that the best way to SOLVE THE PROBLEM OF PCB POLLUTION in the Acushnet River would be to dredge the most contaminated sediments from the river and to incinerate (burn) the sediments in a special incinerator. This incinerator would have been temporarily assembled in the North End of New Bedford. Now, the has been CHOSEN. polluted dredging is under way but the incinerator decision suspended while A NEW CLEAN-UP TECHNOLOGY IS The next nine questions concern the CLEAN-UP of the Acushnet River sediments.
is. Do you think that it is SAFER TO HUMANS, AND BETTER FOR THE ENVIRONMENT, for the EPA to REMOVE (dredge) the most contaminated sediments from the Acushnet River, or to LEAVE them in place?
Do 18. In terms of HEALTH RISKS, what specific aspects of INCINERATION would most concern you? (Please RANK 1 -5, where 1 is the most important, 5 is least important.)  There will always be some degree of trade-off between the need to guarantee public health and safety, and the need to find a cost-effective way of cleaning up the contamination in New Bedford Harbor. The following questions concern the COST OF CLEANING UP THE HARBOR, whether by incineration or any other means. Regardless of the method finally chosen, the cleanup of the harbor will cost many millions of dollars.
27. What do you consider to be the BEST BALANCE between assurance of PUBLIC HEALTH, and COST of cleanup.
clean-up must be as inexpensive as possible     Coverletter for administration of survey to public random sample (Group 2), Portuguese.
Este questionario e para voce poder dar a sua opiniao, da area que o Governo Federal chama o "New Bedford Harbor Superfund Site." Nos esperamos que este questinario no leve mais de que 15 minutes para complitar. Estas preguntas sao para voce dar a sua opiniao, emtam nao a uma resposta que e mais certa de que outra.
Por favor complete todas as preguntos.
Se voce tern alguma pregunta por favor entre en contacto com nos.
Pode  As seguintes preguntas sao dirigidas a o projecto "Superfund Site" no porto de New Bedford. Umas das preguntas faz uso do "Likert Scale" para symbolizar o disposto de valor entre duas situacaos. Marca com um "X." Enclosed are copies of the survey which we discussed by telephone this week, along with a couple of return envelopes. I realize how busy you are; thanks for agreeing to take the time to participate in this study.
If some of the questions seem rather basic, please bear with me; parts of this survey are being administered to a broad cross-section of people, not all of whom are likely to be as familiar with this type of site as you are.
I don't think the survey will take any more than 15 minutes of your time. There are, of course, no right or wrong answers; please read the questions carefully, then answer them in the way that best reflects your personal opinion.
I am not interested in official statements or agency positions, but rather, would like to know how you, personally, perceive this situation .
The information which you provide will be kept completely confidential. The surveys will be destroyed once the results are tabulated. Again, I realize the demands of your schedule, but would be most appreciative if you were able to return the survey within a week or two.
If you have any comments, questions, or complaints about the survey, you may contact myself; . my advisor, Dr. Niels West, 401-792-2596;  Survey questions specific to professional groups (Groups 3 and 4) .

SURVEY --
Concerning the NEW BEDFORD HARBOR SUPERFUND SITE, Upper Acushnet River, New Bedford, Massachusetts.
The following survey should take about 15 minutes to complete. Please read each question carefully, then answer them as completely and accurately as possible. There are no right or wrong answers; please note that I am looking for your personal opinion, not an official agency or departmental position. Your responses will be kept completely confidential .
Thank you for your time .

Date:
Have you lived in the New Bedford area?
If so, for how long?

1.
How °long have you environmental management? years been involved, years -----generally, in 2. In working on a particular site, how often, on average, do you converse with non-professionals stakeholders interested in a site?

3.
Please characterize your role in site management by RANKING the following, where 1 is the most important, 5 is least.