Exploring the Decision Making Process of Prevention Providers in Adopting or Rejecting a Community-Level Approach to Substance Abuse Prevention

Interventions for the prevention of alcohol, tobacco and other drug use (ATOD) have become more sophisticated (Biglan, Mrazek, Carnine, & Flay, 2003; Botvin & Griffin, 2005), partly because the demand for accountability from federal and private funding has increased (Gorman, 2002a, 2002b). Community-level interventions1'2 are multi-component interventions that combine individual and environmental change strategies across multiple settings to prevent dysfunction and promote well-being among population groups in a defined local community (Wandersman and Florin, 2003). A community-level intervention delivered by a community coalition is a model being advocated in the academic literature (Warner, 2000) and increasingly promulgated by federal funding agencies. There is, however, little or no literature on the decision-making process of the community-based organizations and coalitions who must choose whether to adopt or reject this model. This study sought to answer three main questions: 1) to determine how many characteristics3 in the individual and organizational stages of Rogers' innovationdecision process (1995) were used by the community-based organizations and coalitions in their decision making process; 2) to determine which characteristics were most influential in the decision to ~dopt or reject the community-level intervention; and 3) to determine whether the characteristics that influence a decision to adopt differ for organizations representing culturally diverse communities. I Tenns such as "community-based", "comprehensive community", "community coalition", and ~'collaborative partnerships" have also been used to refer to interventions with similar characteristics. Local community is often geographically defined (e.g. , neighborhood or municipality) but may be a ~o~~ity of presumed common interest (e.g., the gay community). This is to be contrasted with an expectation that Roger's stages would be followed in sequence. Rogers himself (1995) states that stages may NOT follow the specific order outlined in his model. For example, in the individual process the stage order might be knowledge-decision-persuasion. Fourteen participants from seven organizations funded through the community-level intervention funding pool and seven organizations funded through an alternative funding pool to implement evidence-based curricula completed a mixedmethod, semi-structured interview between February 21 and April 4, 2008. The questionnaire was designed to capture information regarding the decision-making process of the organization. Participants were first asked a number of non-guiding, open-ended questions before progressing through the remaining sections of the interview which intentionally guided the respondent systematically through select stages in Rogers' individual and organizational innovation-decision process. Findings from the analyses support the application of Rogers' organization innovation-decision process model in combination with specific characteristics from the individual model for understanding community-based organization and coalition funding decisions. The most relevant and influential stages and overarching characteristics from the models were the Matching stage, Characteristics of the Decision-maker, Characteristics of the Innovation, and Communication Behavior. An analysis of organizations serving culturally diverse organizations did not reveal any differences in coded themes, however, the low number of such organizations in this sample may have been a factor. A discussion of the findings and relevant implications are provided as well as a summary of the limitations ofthis study. Acknowledgements I owe a great huge thanks to many people beginning with my family and friends who played a part in helping me complete this project; this is as much mine as it is theirs. Without their support throughout this long process whether offering a suggested revision, hand delivering a form on my behalf, or just checking-in on my progress I probably would not have been able to pull this off. It truly took a community to complete this dissertation! I'd like to send a special thanks to my major professor, Paul Florin, who kept me going and focused. He said it was all part of the job but that did not take away from the feeling that he had gone the extra mile to help me finish which was great since I was 300 miles away! He came through in the clutch and I am sincerely thankful for it. I'd like to thank my committee, Betsy Cooper the chair, John Boulmetis, Al Berman and John Stevenson. John Stevenson particularly provided guidance throughout the entire process and has been a mentor for many years; thank you John!! I'd like to thank my .current employer, the Maine Center for Public Health. They gave me the time and in-kind to collect my data which was absolutely critical. I'd like to thank the participants who gave their time to share their experiences with me. Without their input I wouldn't have had as much to talk about! And most importantly, I would especially like to thank my partner for her patience and support. She took care of our infant twins to give me the time I needed to finish this project which for those who have not raised twins, you'll just have to take my word for it it's an unbelievably huge, huge thing!!!

Fourteen participants from seven organizations funded through the community-level intervention funding pool and seven organizations funded through an alternative funding pool to implement evidence-based curricula completed a mixedmethod, semi-structured interview between February 21 and April 4, 2008. The questionnaire was designed to capture information regarding the decision-making process of the organization. Participants were first asked a number of non-guiding, open-ended questions before progressing through the remaining sections of the interview which intentionally guided the respondent systematically through select stages in Rogers' individual and organizational innovation-decision process.
Findings from the analyses support the application of Rogers' organization innovation-decision process model in combination with specific characteristics from the individual model for understanding community-based organization and coalition funding decisions. The most relevant and influential stages and overarching characteristics from the models were the Matching stage, Characteristics of the Decision-maker, Characteristics of the Innovation, and Communication Behavior. An analysis of organizations serving culturally diverse organizations did not reveal any differences in coded themes, however, the low number of such organizations in this sample may have been a factor.
A discussion of the findings and relevant implications are provided as well as a summary of the limitations ofthis study.  Interventions for the prevention of alcohol, tobacco and other drug use (ATOD) have become more sophisticated (Biglan, Mrazek, Carnine, & Flay, 2003;Botvin & Griffin, 2005), partly because the demand for accountability from federal and private funding has increased (Gorman, 2002a(Gorman, , 2002b. The field has shifted from the delivery of single curricular programs delivered by community-based organizations to the utilization of community-level interventions delivered by community coalitions.

List of Tables
Community-level interventionsl,2 are multi-component interventions that combine individual and environmental change strategies across multiple settings to prevent dysfunction and promote well-being among population groups in a defined local community (Wandersman and Florin, 2003). For example, a community-level intervention for tobacco control might combine a school curriculum for youth to prevent initiation of smoking and a media campaign aimed at reducing parental smoking in the presence of youth (individual change strategies) with policy change efforts advocating a municipal smoking ban for restaurants and increased enforcement of ordinances prohibiting youth access to tobacco.
A community-level intervention delivered by a community coalition is a model being advocated in the academic literature (Warner, 2000) and increasingly promulgated by federal funding agencies. There is, however, little or no literature on I " Terms su~h as "community-based'', "comprehensive community", "community coalition", and 2 collaborative partnerships " have also been used to refer to interventions with similar characteristics.
Local community is often geographically defined (e.g., neighborhood or municipality) but may be a community of presumed common interest (e.g., the gay community).
1 the decision-making process of the community-based organizations and coalitions who must choose whether to adopt or reject this model. "A community coalition is a formal alliance of organizations, groups and agencies that have come together to work for a common goal" (Dluhy, 1990 as cited in Florin, Mitchell, & Stevenson, 1993, p. 417). The use of coalitions to deliver community-level interventions for a "synergistic effect on the whole community" (Florin, Mitchell, & Stevenson, 1993, p. 428;Goodman, Wandersman, Chinman, Imm, & Morrissey, 1996) is not a novel concept. Community coalitions began receiving more attention in the late 1980's but it was only within the past decade that Rhode Island prevention providers were given a fiscally-supported opportunity to utilize the coalition-based, community-level intervention approach.
The State of Rhode Island was awarded a State Incentive Grant (SIG) by the Center for Substance Abuse Prevention at the Substance Abuse and Mental Health Services Administration, DHHS. The SIG funds were divided into several pools, one of which required the implementation of a community-level intervention by an established community coalition. This contract was extremely demanding and challenging, asking Rhode Island prevention providers to deliver services in a manner in which they had little or no experience. Indeed, creating and sustaining coalitions is difficult (Libby & Austin, 2002;Lackey, Welnetz, & Balistrieri, 2000) and using them as a delivery mechanism for prevention is a "complex and difficult model" (Florin, Mitchell, & Stevenson, 1993, p. 417). Today, despite the established credibility and acceptance of the community-level intervention approach, prevention service providers must still decide whether to adopt this approach to substance abuse 2 prevention and the question arises, "Are prevention providers ready to adopt this approach?" This dissertation addressed this question by assessing the "readiness" of prevention providers to adopt this model now being strongly promulgated from the federal level. Furthermore, it sought to identify which characteristics most influenced the decision to adopt. This can potentially provide guidance for training and technical assistance services to increase the probability of adoption of this new model. Finally, it tried to identify characteristics most important in the decisions of providers who service minority communities. This has the potential to enable more culturally sensitive approaches to promoting adoption.

Rogers' Innovation-Decision Process as a Conceptual Framework: General Description
This study employed Rogers' Innovation-Decision Process as a conceptual framework for analyzing the decision-making process prevention providers used when deciding to adopt or reject a community-level approach to substance abuse prevention. Rogers (1995) describes the "innovation-decision process" as an "information seeking and information-processing ll:ctivity" (p. 165). He further states that this is a social process that occurs over time and distinguishes it from other decision-making processes in that the decision is about something new and that there is an uncertainty involved in adopting an innovation as a "new alternative to those [other innovations] previously in existence" (Rogers, 1995, p. 161 ). Rogers outlines separate innovationdecision processes for individuals and organizations. Both are, however, essentially information seeking and processing activities. The concepts explained within the individual process contribute to the organization process (Rogers, 1995).
The innovation-decision process for an individual is defined by Rogers (1995) as "the process through which an individual (or other decision-making unit) passes (1) from first knowledge of an innovation, (2) to forming an attitude toward the innovation, (3) to a decision to adopt or reject, (4) to implementation of the new idea, and (5) to confirmation of this decision" (p. 161). This study focused only on the initial stages of the individual decision-making process (i.e., Knowledge, Persuasion and Decision), shown in Figure 1.     The innovation-decision process in an organization also consists of five stages similar to the individual process but tailored to an organizational structure. The organization innovation-decision process consists of two subprocesses: "initiation" and "implementation" (Rogers, 2003, p. 422). This study focused only the "initiation" subprocess which contains the first two stages: (1) Agenda Setting and (2) Matching shown in Figure 2. It is in the transition from the initiation to implementation subprocess that a Decision is made. The innovation is modified and reinvented to fit the organization, and organizational structures are altered IV. CLARIFYING ..

' ' '
The relationship between the organization and the innovation is defined more clearly V ROUTINIZING .. ' ' ' ' ' The innovation becomes an ongoing element in the organization's activities, and loses its identity

Description of Stages Investigated in This Study
Individual Innovation-Decision Process: Prior Conditions 5 "The innovation-decision process is essentially an information-seeking and information-processing activity in which an individual is motivated to reduce uncertainty about the advantages and disadvantages of an innovation" (Rogers, 2003, p. 172). As a precursor, Hassinger (1959, as cited in Rogers, 2003 states that decision-makers must experience a felt need or problem to be open to messages about an innovation. This perceived need should be sufficient to mobilize an individual, organization or community to search for a new solution (Price, 2000). There are, however, occasions when knowledge of an innovation precedes the felt experience of a need.
Other precursors include previous expenence with the innovation, or innovations in general; innovativeness of the potential user; and norms of the social systems within which the individual travels. Successful adoption and implementation of innovations on previous occasions will have a positive impact on future decisions to adopt innovations and on the success of future implementation.

Individual Innovation-Decision Process: Communication 6
Communication channels influence the innovation-decision process at every stage. Rogers (2003) defines communication as "a process in which participants create and share information with one another in order to reach a mutual understanding" (p. 5). This process involves those who have knowledge of and experience with the innovation, those that do not, and the communication channels (interpersonal, mass media) between them. The communication channels are an important aspect in that "most individuals evaluate an innovation not on the basis of scientific research by experts but through the subjective evaluation of near peers who have adopted the innovation" (p. 36). The communication process is also differentiated by the fact that it is about some new idea, practice or object.
The specific qualities of the communication process can influence the potential user's decision to adopt or reject the innovation (Dearing, 2004;Klein, 2000). There are a number of key strategies that act as facilitators or barriers to adoption (Backer, David, & Soucy, 1995). Morrissey, Wandersman, Seybolt, Nation, Crusto, & Davino (1997) outline several barriers to adoption including differing funding priorities, resource constraints, systems-level barriers, community readiness, and differing theoretical orientations between the developers (social scientists) and the users (practitioners). They define barriers as "factors which prevent the dissemination of information or which make it difficult for practitioners" to adopt appropriate innovations (p. 373).

Individual Innovation-Decision Process: Knowledge
The individual innovation-decision process begins with the knowledge stage (Rogers, 1995, p. 162) in which a potential decision-maker first becomes aware of an innovation before seeking information about that innovation. These decision-makers may learn of an innovation's existence from change agents or through colleagues (social networks). Others who experience a perceived need or problem may proactively search for a potential solution using whatever resources are available and accessible. Rogers (1995) outlines three types of information gathering during the knowledge stage. The first type is "awareness-knowledge" (e.g., "What is the innovation?") or knowing that an innovation exists (p. 165). Rogers (2003) states that, "at any given point in time, many potential adopters are aware of a new idea, but are not yet motivated to try it" (p. 213). In the current study, CBOs and coalitions were given a short list of evidence-based programs which were approved for implementation under the SIG RFP. They were also supplied with resource links to learn more about the approved evidence-based programs as well as the environmental strategies. "How-to knowledge" (e.g., "How does it work?") represents gathering knowledge on how to use the innovation (Rogers, 1995, p. 165). It was the responsibility of the CBOs and coalitions to learn how the evidence-based programs and environmental strategies worked in order to form an opinion about whether or not it would fit within their context. The third type is "principles-knowledge" (e.g., "Why does it work?") which is gathering info on the underlying function of the innovation (Rogers, 1995, p. 165). This author is unaware of CBO or coalition efforts to gain principles-knowledge. Also, the knowledge stage can be influenced by the characteristics of the decision-maker including "socioeconomic characteristics, personality variables, and communication behavior" (Rogers, 1995, p. 163).

Individual Innovation-Decision Process: Persuasion
In the Persuasion stage the potential user forms a positive (favorable) or negative (unfavorable) attitude toward the innovation (Rogers, 2003). Rogers (1995) defines the decision-makers' actions in this stage as seeking "innovation-evaluation information to reduce uncertainty about an innovation's expected consequences" (p. 168). The potential user is actively seeking information about the innovation's advantages and disadvantages as they relate to the problem or need. The potential user must then decide how to interpret the information. These decision-makers will often share their perceptions and initial attitude about an innovation with their peers in an attempt to reinforce their beliefs (Rogers, 1995). It is important to note that the decision-maker's attitude does not always align with the final decision (Rogers, 1995).
It is during the persuasion stage that potential users consider the characteristics of the innovation. The innovation is the actual "idea, practice, or object that is perceived as new by an individual or other unit of adoption" (Rogers, 2003 p. 12). In this study the innovation is the community-level intervention approach implemented through a coalitional structure. Potential users are influenced in their decision-making by the characteristics of the innovation, what Rogers defined as the Theory of Perceived Attributes (Backer, Liberman, & Kuehnel, 1986: Klein, 2000Mayer & Davidson II, 2000;Rogers, 1995). There are five main attributes identified by Rogers: 1. Relative Advantage; 2. Compatibility; 3. Complexity; 4. Trialability; and 5.
These five attributes explain most of the variance m decisions to adopt (Rogers, 2003). A study by Keams (1992, as cited in Rogers, 2003 investigated 25 perceived attributes and found that they explained 27% of the variance. When the aforementioned five attributes were removed from the 25 and analyzed separately, they explained 26% of the variance (Rogers, 2003).
Knowing the characteristics considered by potential users, including the relative importance of each characteristic, has implications for the dissemination of coalition, community-level interventions and the development of requests for proposals at the state level. Description "The degree to which an innovation is perceived as being better than the idea it supercedes" (Rogers, 2003, p. 229). This attribute contains subdimensions such as conferring social status, providing a cost/economic benefit, saving resources, and decreasing discomfort. This attribute has the most supporting evidence (Rogers 2003).
"The degree to which an innovation is perceived as consistent with the existing values, past experiences, and needs of potential adopters" 9 Complexity Trialability Observability (Rogers, 2003 , p. 240). This attribute is intercorrelated with relative advantage but Rogers' states that it is conceptually different.
"The degree to which an innovation is perceived as relatively difficult to understand and use" (Rogers, 2003, p. 257). The opposite of complexity is simplicity, or how easy an innovation is perceived to use.
"The degree to which an innovation may be experimented with on a limited basis" (Rogers, 2003, p. 258). Rogers (2003) states that most potential users test the innovation on a small scale for a limited time, however, organizations were responding to an RFP and if awarded funding, would enter into a contract with the state to provide services.
With no opportunity to try the innovation this attribute is not applicable.
"The degree to which the results of an innovation are visible to others" (Rogers, 2003 , p. 258

Individual Innovation-Decision Process: Decision
The decision-maker is ultimately faced with making the decision to adopt or reject the innovation. This decision can be made by one individual or by the entire social system (see Brink, Basen-Engquist, O'Hara-Tompkins, Parcel, Gottlieb, & Lovato, 1995;Parcel, 1995;Parcel, O'Hara, Harrist, Basen-Engquist, McCormick, Gottlieb, & Eriksen, 1995;Parcel, Taylor, Brink, Gottlieb, Engquist, O'Hara, & Eriksen, 1989 for exemplars). The social system is defined as "a set of interrelated units that are engaged in joint problem solving to accomplish a common goal" (Rogers, 2003, p. 23). The communication process occurs within the social system l ·ndividuals with shared characteristics in similar networks (e.g., between between members of different CBOs and coalitions) and with more difficulty between individuals who are not similar and have different social networks (e.g., between developers, State Departments, and CBOs I coalitions).
"The social and communication structure of a system facilitates or impedes the diffusion of innovations in the system" (Rogers, 2003, p. 37). The system often consists of, for example, individuals (opinion leaders, change agents), and/or organizational characteristics (social structure, norms) that influence the adoption or rejection of an innovation.
There are three different types of innovation-decisions: optional, collective, and authority innovation-decision (Rogers, 2003, p. 403). An optional innovationdecision is when one member of a social system makes a decision independent of other members. Collective innovation-decisions are made by consensus among members of a social system. Authority innovation-decisions are when the few individuals in a system who posses the power and knowledge make the decision. The latter two are more relevant for the organization decision process.
The decision to reject an innovation can occur at any time during the innovation-decision process. There is both "active" and "passive" rejection (Rogers, 2003, p. 178). Active rejection describes decision-makers who made the decision not to adopt. Passive rejection describes decision-makers who forgot or dropped out of the innovation-decision process before actually making a decision.
Organization Innovation-Decision Process: Organizational Characteristics Rogers (2003) describes an organization as "a stable system of individuals who work together to achieve common goals through a hierarchy of ranks and a division of labor" (p. 404). The organizational structure contains: 1. predetermined goals (purpose, mission); 2. prescribed roles; 3. an authority structure; 4. rules and regulations (by-laws); and 5. informal patterns (practices, norms, social relationships).
Some of the characteristics of an organization that influence adoption include: centralization, complexity, formalization, interconnectedness, organizational slack, size, and system openness (Rogers, 2003). A champion, or an opponent, can also influence an organization' s innovation decision. The champion is equivalent to an opinion leader. Table 2 contains a list of the organizational characteristics that influence the decision to adopt or reject an innovation.
These characteristics have been researched in several hundred studies of organizational innovativeness and results indicate that there is a "low" correlation between these independent variables and organizational innovativeness (Rogers, 1995, p.381). A study by Meyer andGoes (1988, as cited in Rogers, 2003) found that the characteristics of the innovations (i.e., perceived attributes) explained 40% of the variance while characteristics of the adopting organization only explained 11 %. One possible explanation for the low variance might be that different characteristics are important at different points in the stage model (Rogers, 2003). For example, "low centralization, high organizational complexity and low formalization facilitate innovation in the initiation subprocess, but impede implementation" (p. 433). Definition "The degree to which power and control in a system are concentrated in the hands of a relatively few individuals" (Rogers, 2003, p. 412).
"The degree to which an organization's members possess a relatively high level of knowledge and expertise, usually measured by the members' range of occupational specialties and their degree of professionalism" (Rogers, 2003, p. 412). This is expressed by formal training.
"The degree to which an organization emphasizes its members' following rules and procedures" (Rogers, 2003, p. 412).
"The degree to which the units in a social system are linked by interpersonal networks" (Rogers, 2003, p. 412).
Definition "The degree to which uncommitted resources are available to an organization" (Rogers, 2003, p. 412).
This is the best predictor of organizational innovativeness (Rogers, 2003). This characteristic encompasses the number of staff, size of budget, size of community they serve, prestige and social influence System openness (positive) of the decision-maker among their peers, credentials, staff technical expertise, and cosmopoliteness as evidenced by such things as how many boards and committees one belongs to outside the organization.
"The degree to which the members of a system are linked to other individuals who are external to the system" (Rogers, 2003, p. 408) Note. Positive or negative in parentheses indicates whether that characteristic would be a facilitator or barrier as it relates to the definition.

Organization Innovation-Decision Process: Agenda Setting
Agenda setting "occurs when a general organizational problem is defined that creates a perceived need for an innovation" (Rogers, 2003, p. 422). In this stage the organization identifies and defines a problem or need. The organization then prioritizes the problems or needs before actively searching for a solution. Rogers (1995) does state that an organization might encounter an innovation without having initially identified a problem or need. This stage for organizations is similar to the knowledge stage for individuals.

Organization Innovation-Decision Process: Matching
The matching stage for the organization shares the characteristics of the persuasion stage from the individual innovation-decision process with an additional layer of considering the contextual fit of the innovation for the organization.
Matching involves an attempt to fit the innovation with the organization's need, context and values. In this stage, the organization is actively trying to determine how well the innovation would address its need while also determining how seamlessly the mnovation could be implemented within the organization's structure and climate. If there is a good fit then adoption is likely to occur while a mismatch is most likely to lead to rejection.

Questions of the Current Study
With the emergence of the community coalition as an intervention platform for implementing a community-level approach over the past decade, it is important to understand the perceptions of potential users and illuminate their decision-making processes. This study sought to answer three main questions. The first research question was to determine how many characteristics 7 in the individual and organizational stages of Rogers' innovation-decision process were used by the community-based organizations and coalitions in their decision making process. This question was answered by determining which of the characteristics were mentioned by decision-makers by applying the codebook to the transcripts in two parts. The first analysis reviewed the participant's responses to two questions that did not guide them through the characteristics. The second analysis was applied to the entire transcript which included the guided questions.
The second research question was to determine which characteristics were most influential in the decision to adopt or reject the community-level intervention.
This question was answered in two parts by determining which characteristics were mentioned most often during the interviews. The first analysis reviewed participant responses to one interview question about the factors participants felt were most unportant in influencing their organizations' decision to apply or not for the community-level intervention funding. This interview question was designed to ask ~This is ~o be contrasted with an expectation that Roger's stages would be followed in sequence.
ogers h~self (1995) states that stages may NOT follow the specific order outlined in his model. For example, m the individual process the stage order might be knowledge-decision-persuasion. 15 about the decision process without providing any particular prompt to the respondent; thus allowing the respondent to offer the underlying reasons for their decisions with minimal bias introduced by the interviewer and interview process. The second part used an alternate approach for considering the most influential characteristics by identifying the characteristics most often mentioned by respondents. The more frequently a characteristic was mentioned, the more important that characteristic was to the decision-making process.
The third question to be answered by this study was to determine whether the characteristics that influence a decision to adopt differ for organizations representing culturally diverse communities. The third research question was answered by comparing responses to one non-guided interview question from organizations serving culturally diverse communities with other organizations not serving culturally diverse populations.

Setting
The State of Rhode Island was awarded a State Incentive Grant (SIG) in 2001.
The SIG established a funding pool for community-level interventions. Community-Ievel intervention was specifically defined as implementing evidence-based programs at both the individual and family levels, in conjunction with environmental strategies (two each for policy, media advocacy, and enforcement) at the community level. The community-level intervention was to be implemented only by a community coalition that had been in operation for a minimum of one year.

Participants
Participants were fourteen Executive Directors and Coordinators from community-based organizations (CBOs) and community coalitions who were involved in the final decision of whether or not to apply for the community-level intervention funding. 8 Seven represented coalitions who applied (Group 1) and seven represented CBOs or coalitions who did not apply (Group 2).
The participants in Group 1 represented 87 .5% of the coalitions who applied for and received funding for the community-intervention (7 of 8, 9 with one declining to participate). Participants in Group i1° represented 50.0% of the CBOs or coalitions who chose not to apply for the community-level intervention funding (7 of a total 14 0 . . II) rgan1zations , but instead chose to apply for an alternative funding pool to 8 Tur .
ee mterviews were removed from the Group 2 sample to maintain consistency within the group.
~e thr. ee that were removed were funded to implement programs that their organization had developed. 9 ey did not have to adopt an innovation. 10 Two organizations applied for both the community-level intervention and alternative funding pool. fil c;oup 2 originally contained additional organizations who had not filed a letter of intent or who had . e s.uch a letter but did not apply. However, contact with seven such organizations yielded no mterviews due to personnel turnover or limited recall of what was considered a minor event.
implement evidence-based curricula. Twelve of the 14 organizations were funded; all ·n Group 2 were awarded funding. seven 1 The fourteen participants, ten females and four males, identified their roles as Vice President (n=l), Executive Director (n = 4), Managers of specific departments (n ::::: 4), Coordinators (n = 3) or Assistant Managers (n =2). Nine participants identified as community-based organizations (in operation for an average of 50 years), four as Substance Abuse Prevention Task Forces serving specific Rhode Island municipalities (self-identified as coalitions), all operating for twenty years, and one self-identified as "other." Only four organizations (one from Group 1 and two from Group 2) identified themselves as working specifically with culturally diverse populations. The additional organization was one of the three that had been removed from Group 2 for being funded to implement a program that they had developed locally.

Procedure
Prior to contacting the selected participants this author contacted the Prevention and Planning Unit Administrator at DBH to notify him of the study and discuss the potential relevancy of the results for the state department. The Administrator supported the effort and agreed to email potential participants. The organizations selected for participation were telephoned by a student intern in early February to verify and update contact and mailing information. The initial verification calls uncovered that some identified contact persons were no longer with the organization. Potential participants were sent invitational letters (Appendix A) informing them of the study, protections for them as participants, and the author's t . n to call to schedule an interview in the following one to two weeks. The inten 10 also mentioned that the participants could not be compensated for their time. letters The student intern made the initial telephone calls on February 15 to request participation and to schedule a one hour telephone interview with the author. One week later the author began making the calls to schedule the interviews. The initial conversation was scripted on the Interview Guide & Questionnaire (Appendix B). The script introduced the study and inquired into whether the participant had in fact received the invitational letter. There were a number of individuals that did not receive the invitational letter as they had changed employment or could not remember seeing the letter. These individuals were faxed or emailed the letter between the time of the initial contact and the scheduled interview. At the beginning of the interview the participant was asked again if they had received the letter.
After inquiring about the invitational letter, the individual was asked if they were interested in participating and whether they could schedule a one-hour meeting.
Issues with individuals no longer present at the organization were discussed and resolved during the initial contact by either identifying another appropriate contact within the organization or providing a last known place of employment to give the author an opportunity to try and find the individual's new contact information.
The first scheduled semi-structured telephone interviews were completed on February 21, 2008. There was information in the letter and on the Interview Guide & Questionnaire that notified participants of their rights as human subjects and that the interviews were confidential despite there being no expected risk to the participants.
Participants had the opportunity to verbally consent at the beginning of the process d Id decline to participate at any time. They were also notified that the interview an cou would be recorded and destroyed as soon as it was transcribed by the student intern and verified by the author. Participants were also asked if they wanted to receive a copy of the dissertation once it was completed. The interviews, averaging approximately 40 minutes in length, were completed between February 21 and April 4, 2008.

Measures
Most of the studies on diffusion have employed quantitative methods of inquiry. Rogers' (2003) recommendation for better understanding the innovationdecision process is to use more qualitative research methods such as in-depth personal interviews. A mixed-method survey (Appendix B) was thus created by the author.

20
The Interview Guide & Questionnaire was not pilot tested, however, during the course of the interviews modifications were made to specific questions and section descriptions. These are described in the Interview Guide and Questionnaire found in Appendix B as technical notes. One change emerged during the first interview with a Group 2 participant that should be noted here. A follow-up question was added to 2 .9b in response to the first interview with a Group 2 participant. The questions in sections two and three were focused on the community-level intervention approach to determine the reasons why Group 2 participants did not pursue that funding. When asked question 2.9b, the respondent provided the reasons why they had applied for the alternate funding. When asked directly about the community-level intervention, the respondent did provide information as to why they had not pursued that funding. This

Design and Analyses
This study employed a multiple case study design using a mixed-method, retrospective interview. This multiple case study design was the most appropriate as the main purpose of this study was to test the various aspects of the innovationdecision process and not necessarily individual or group differences, "generalization of the results from ... case study design is made to theory, not populations" (Yin, 1993, p. 79). A quasi-experimental design or even mixed design was ruled out as there was no random assignment or comparison group, and while subjects are grouped, the N is too low and would not provide for sufficient power in conducting group comparisons.
II Th~se ~uestions were based on the research of Miller (2001) who identified these specific organizational characteristics as most relevant in relation to adoption in the HIV prevention field.
22 f a descriptive case study design allows for a phenomenon to be studied The use o "within its context" which was important for the purpose of this study (Yin, 1993, p. 5 A codebook was developed usmg Rogers' individual and organizational decision-making model and characteristics, as well as other characteristics from the research literature that had not been defined by Rogers (Appendix C). The codebook was organized by stages (e.g., Knowledge, Persuasion, Matching, Agenda Setting, Decision) and overarching characteristics (e.g., characteristics of the decision-maker, characteristics of the innovation, prior conditions). The stages and overarching characteristics were the themes. With the exception of the matching and agenda setting stages, the other stages and overarching characteristics served as an "umbrella" for a collection of different (sub-)characteristics which were specifically used during the coding of the transcripts. These sub-characteristics were the subthemes. Since the matching and agenda setting stages did not consist of any identified sub-themes, they were specifically applied when coding text. Coded text included any phrase(s) or sentence(s) that aligned with the definition of a specific theme or subtheme as outlined in the codebook. There were a few instances when single words were coded. These were often yes or no responses to interviewer questions or probes.
The recorded interviews were transcribed onto a Rich Text Format document by the student intern and then the transcription was verified by the author. The codebook was first applied by hand to the transcripts to refine the coding definitions and to enhance consistency in coding participant responses. The transcripts were then llllported into NVivo and coded within the software program using the revised 23 bo k after which a search was conducted to review each code separately, code 0 ·ng similarly coded text to ensure consistency. Revisions were made as com pan necessary.
Classical content analysis (Ryan & Bernard, 2000) of coded text was used to organize the data for further analysis. This approach was appropriate since the codes (i.e., themes) had already been identified and described in the scientific literature. A matrix of coded themes by participant was generated. Each CBO or coalition was the unit of analysis or case; together forming multiple cases.
Pattern matching was used to analyze the data by matching the observed patterns of decision making (i.e., number of times a theme had been coded by participants within groups) to that of Rogers' individual and organizational innovation-decision process. This approach compares the patterns in an attempt to support or refute the innovation-decision process and its characteristics. Taken within their group, the identified patterns by case serve as their own replication.
In reviewing the coded text, specific concepts related to each theme were sought. These concepts were consistent groupings of similar topics (Ryan & Bernard, 2000) used by respondents to characterize or describe their decision-making process.
For example, relative advantage was coded as a theme. Relative advantage is how much a new innovation is perceived as better (or not) than what currently exists (Rogers, 2003). A concept within the relative advantage theme for some respondents was the perceived opportunity afforded by the funding to expand on existing services.
The quantitative information from the Interview Guide & Questionnaire were coded then entered into a Microsoft Excel Spreadsheet by the student intern and 24 cleaned by the author. This spreadsheet was converted into a Rich Text Format document then imported into NVivo. The quantitative data were imported as attributes and linked to the respective participant.

Research Question 1
The first research question was to determine the extent to which the characteristics and stages from Rogers' innovation-decision processes were used by the community-based organizations and coalitions in their decision making process. A review of the coded text and content analysis was conducted on the transcripts of the seven organizations that were funded to implement the community-level intervention (Group 1) and the seven organizations that were funded to implement one evidencebased program (Group 2).
The analysis was conducted in two parts using the codebook (Appendix C) that reflected Rogers' concepts from both the individual and organizational decisionmaking model. The first analysis reviewed the participant's responses to two questions that did not guide them through the characteristics. The second analysis was applied to the entire transcript which included the guided questions.

Non-guided Analysis for Research Question 1
Some interview questions were designed to elicit responses from participant's experiences without introducing Rogers' characteristics. By not guiding participants through the characteristics, they would share the more salient aspects of their decisionmaking process. Specifically, two questions were used in this fust analysis: 1.) the reasons why (or why not) the organization applied for the community-level intervention approach; and 2.) what things were considered in the process of making the decision.

26
Twenty-one different stages and characteristics were mentioned at least once by respondents. Eight stages and characteristics were mentioned most frequently.
These are presented in Table 3.

Matching.
Matching the different grant requirements, including perceived advantages and disadvantages, with various characteristics of the organization was the most mentioned l t ·on by respondents for why they applied as they did. These non-guided exp ana I resPonses strongly support Rogers' organizational model, specifically the matching The findings also support the characteristics from both the individual and stage.
aill ·zation decision-process models, and the importance for organizations of org matching innovations with their organizational structure, capacity, values, and the needs and values of their partners and the community they serve. Most often there were different combinations of characteristics mentioned by the respondents in how they matched, but the results that follow were clustered to better present commonalities.
Matching the characteristics of the decision-maker with the grant requirements was cited most often by respondents (22 coded texts). Comments reflected a match between the structure, capacity, resources, staff expertise, and relationships in deciding which funding pool to apply. For example, a respondent from Group 1 stated: Well I think our coalition has built an infrastructure to be able to support those kinds of activities in the community and our role, we are seen as kind of the hub of the wheel for substance abuse prevention activities. So we saw it was a perfect match for how we were structured and what the grant requirements were.
Some respondents from Group 2 would state, however, that they did not have the structure, capacity, staff expertise and relationships to apply for the communitylevel intervention but did possess these characteristics for the other funding pool. 1 think it was a determination of capacity at the moment in that we felt that our strength at that time in relation to the grant that was out there, was appropriate for the one that we applied for. We didn't have necessarily a community coalition specifically designated around substance abuse prevention in place at the time. And I think our sense was there were other groups that did who would have a distinct advantage over us in terms of that funding stream. It was more of a strategic determination.
Having previous experience with an intervention (seven coded texts) was mentioned as a reason why the funding requirements were a good match for the organization.
Associated with already implementing the same or a similar intervention was the perceived opportunity to expand on that intervention (i.e., relative advantage; four coded texts). One respondent from Group 1 said: We were already doing the work but we could expand the scope of work. And we were in a good position to do that because we already had the expertise, we had the relationship, and it just seemed like a win-win.
Four respondents in Grpup 2 applied for their funding pool because they were already doing a similar intervention and felt they could expand that work; that expanding services by building on an existing intervention that fit with the other funding pool was easier or more feasible.
Because we already serviced the population, one of the target populations that was identified, and we thought that we were already a provider of different types of services in the community and that we had access to that population in 29 treating them in different needs that we would be able to translate that easily to providing another community-based service.
Another respondent from Group 2 stated: Well we had the evidence-based program that we had been interested in replicating, was a very close replication of what we were currently doing in the public schools in Rhode Island, so it felt it was a natural extension of our services to a different population and we saw that as an opportunity to expand on what we were currently offering without stretching our current resources or going off in a completely different direction.
Two other characteristics worth mentioning were matching with organization mission and goals, 12 and with identified needs of the organization or community (four coded texts). Six respondents mentioned matching their organization's mission and goals with the different funding requirements (seven coded texts). It was also mentioned at times as the "direction" the organization may or may not have been heading in, " ... we really didn't consider applying for the coalition money, ever seriously, looked at it and said .no that's not the direction we're going in. <Interviewer: And by direction you're going in, meaning?> Becoming a coalition." Other characteristics.
The remaining coded texts that were not directly linked with matching were of comments about having staff expertise, having identified a need, the financial benefits of the RFP, and time constraints as characteristics that influenced the decision-making process.
Time constraints was not a characteristic original to Rogers; it emerged as a d uring the coding of the transcripts. Three respondents from Group 2 theme t . ed insufficient time to develop a coalition as the primary reason for not men 1on applying for the community-level intervention: "We thought it was the time frame, it was hard to put coalitions together and develop working relationships for what we wanted to do for this grant application." Another respondent stated: "We didn't feel that we had enough time to be able to formulate a coalition with other providers of services in our area." The final respondent said: "Because pulling the coalition together in the amount of time we had, to get the application done was impossible."

Guided Analysis for Research Question I
The second analysis was conducted using the entire transcript of non-guided and guided questions. This analysis expanded on the first to investigate more generally the support for Rogers' stages and overarching characteristics. The results indicated that a combination of Rogers' individual and organizational decision-making characteristics was involved in the decision-making process thus supporting the different stages and overarching characteristics of the model. Table 4 contains the number of coded texts for each stage and characteristic by interview.

Matching.
The most mentioned decision-making characteristic pertained to matching organization and innovation thus further supporting Rogers' Matching stage. Through a communication process organizations discussed the fit of the innovation (i.e., characteristics of the innovation) with their structure, values, needs, partners and population to be served in deciding to apply for the community-level intervention or . 1 idence-based program funding pool.
Previous experience influenced stng e ev · ns about match. Respondents also considered the feasibility of adopting in discuss to terms of resources including financial.  aggregate of its subthemes as defined in the codebook. The number of subthemes differs theme 1s an h th me thus increasing or decreasing the opportunity for a coded text. Group 2 was asked more foreac e . estions which would also provide an opportunity for more coded texts. There are also coded probmg qu texts that reflect 'no' answers on questions such as for contacting developers or opinion leaders which would be in the communication behavior theme.
Respondents mentioned how the requirements of the RFP matched their infrastructure, "perfect match for how we were structured and what the grants requirements were." Another respondent said, " . . .it fit fine, it seemed like a really logical next step for us to be making as an organization." There was also mention of having capacity to implement the intervention( s) which referred to trained staff and for Group 1, having relationships with community partners or having the capacity to develop those relationships.
Group 2 respondents said the requirements of the community-level intervention were too difficult and beyond the scope of their capacity. Structurally they were not coalitions so the single evidence-based program was a more "realistic" and "manageable" fit for the type of work they did. They did not have the staff and other resources to effectively implement the community-level intervention but did have the structure and capacity to more easily add one evidence-based program or build on existing services and previous work with a single intervention. The major obstacle was the requirement for t}ie community-level intervention to be delivered by a coalition. When asked about the influence of each requirement on their decision to not apply, the organizations stated that the major factor was the coalition requirement because they were not a coalition. The multiple evidence-based programs and 33 . runental strategies were not really considered because of the major structural enviro difference.
Respondents mentioned that the funding for the interventions fit with their aru ·zation's values, goals, objectives, and mission. For one organization, coalitions org delivering community-level interventions "fit perfectly because this group had been a coalition and had worked together in the past so they saw it as totally consistent with their thinking." The interventions also fit with the needs of the organization and those they partnered with, the population they served, and the current and previous implementation of interventions and experiences. Some respondents had already identified the intervention that would be the best match or contacted program developers to discuss it. One coalition had spent the previous four years identifying needs. When the "SIG funding came along ... it seemed like a no-brainer to apply for it." There was one respondent who described a compatibility issue between the population served and the menu of interventions. In terms of matching, the interventions did not fit but the organization needed money to sustain itself. The respondent mentioned that years later "in hindsight" s/he saw the value of the approach. The compatibility issue was described as such: ... we were very much aware of certain conditions that exist in the community, but again we don't deal with just one or two conditions, and then when you're dealing face to face with a family, how do you differentiate, I can't say to a family, I can only talk to you about tobacco cessation right now, you cannot talk to me about your homeless issue or your HIV issue. . . and so for us as a grass-roots organization, and that was the thing we struggled with, we were not a, this isn't a classroom environment, we are a community center, we're a community driven organization and so we rise up to meet the needs of what's happening in the community, we don't tell them what we think they need. And some of these science-based programs do that. And we weren't very successful, we were not very successful, I can say that.
For respondents in Group 2, they differed on whether the community-level intervention fit with their organization's values. Within the group they did not differ in stating that the intervention did not fit "with our physical abilities" and it was a different type of work then the organization had done, "not a direction we're going in." Funding also influenced the matching process. Organizations needed to sustain services and staff. One respondent said they matched with the funding they thought they were most likely to receive which happened to be a single evidencehased program. The funding w:as seen as providing a financial benefit.

Characteristics of the decision-maker.
The second most mentioned decision-making characteristic after matching was characteristics of the decision-maker. The specific characteristics that were mentioned were: 1) organization structure and capacity; 2) organization system openness; 3) organization complexity; 4) readiness; 5) organization size; and 6) organization centralization. The concepts in this thematic area differed between Group 1 and 2; consistent with their decisions to adopt and reject, respectively, the community-level t ion approach. The concepts that will be presented are for the organization interven strocture and capacity, system openness and organization complexity subthemes. The other characteristics of the decision-maker were mentioned during the interview process but no concepts emerged. Quantitative information from specific questions about readiness and organization size is presented at the end of this section.
Organization structure and capacity was not as salient an issue for those in Group 1 who were more focused on their relationships with other community partners.
In response to pursuing funding for the community-level intervention, which required a coalition for service delivery, respondents stated that they were coalitions and had the infrastructure to reach the community effectively. One respondent stated, "The community-level one, we knew we could reach the community on a broader level." Another respondent stated, " ... our coalition has built an infrastructure to be able to support those kinds of activities in the community." And another commented, "So I think really we felt like we had the capacity, that's the key, I think that's the key thing." Organization structure and capacity was a more salient issue for those in Group 2 who did not pursue funding for the community-level intervention. Respondents stated that their organization was not a coalition, was not linked to a coalition, and did not have time to develop or become part of a coalition. Five Group 2 pc:µticipants said that the community-level intervention was beyond the scope of their capacity, " be ··· yond the scope of what we felt we could successfully do." On:~,'respondent stated: "We just don't work in the community in that particular way so we never really entertained it seriously." Another respondent said, Because we aren't a coalition and while we partner with the local substance abuse coalitions in the state and work closely with them, we aren't a part of that coalition, or our organization is not a member of a coalition.
For Group 2 respondents, the criteria they used in deciding to not go for the community-level intervention funding were mostly the same criteria used in deciding to go for the other funding pool. Respondents perceived the alternate funding pool as a better fit with their capacity in that it was "more realistic," "manageable," and "more workable." It was also a better fit because it provided an opportunity to build upon existing experiences and resources. One respondent commented, "We knew we had the capacity with the other one. And we were building upon successful programs that were in place at the sites that we were going to deliver the model." Another respondent stated: Well, the community based part of the SIG, funding stream of the SIG, really seemed something we were already positioned to take advantage of. It seemed a natural part of a continuum for us really in our decision to replicate the evidence based prograpi that we selected, was really just moving into a different setting, and not even a very different setting... and using the same program, so there was not, and again, given that it wasn't a whole lot of money to gear up with a lot of staff, we needed to look at something we could capitalize on our existing resources with and we also, the decision was, we really wanted to do something we know we could do well and had some track record with.
Other comments included: "I guess we felt that it was a more realistic thing for ·mplement the single program adaptation rather than pull together a coalition." us to i And, I think it was a determination of capacity at the moment in that we felt that our strength at that time in relation to the grant that was out there was appropriate for the one that we applied for. We didn't have necessarily a community coalition specifically designated around substance abuse prevention in place at the time. And I think our sense was there were other groups that did who would have a distinct advantage over us in terms of that funding stream. It was more of a strategic determination.
Group 2 organizations could have and perhaps did try to connect with local coalitions to form a partnership; however, most respondents did not mention seeking to establish a relationship. They stated that structurally they had the capacity to implement the single evidence-based program and not the community-level intervention. Two possible explanations were identified from three comments made by Group 2 respondents: being time constrained and not wanting to share the funding award.
Since we had not been working with a coalition as part of a coalition up until then, it seemed a timely, it just seemed time consuming a process, we didn't have the resources at that time to invest and it was really by the time the RFP came out, there certainly wasn't time, I think, really you had to have been building that coalition and making some decisions prior to the grant coming out, and there was enough lead time for coalitions to know, I mean there was a 38 lot of lead time for coalitions to know this was coming, but we just don't work in the community in that particular way so we never really entertained it seriously.
Another comment about time constraints was, It was only a very brief discussion about which funding pool we would apply for, it became evident to us pretty early on that we currently weren't part of a community coalition group and that we didn't think there would be the time necessary for us to develop that and become part of that with other service providers in our area.
A couple of comments about maximizing the funding award were, "We just felt like we could keep it largely in house." And, "Just the difficulty in forming those coalitions and then the funding gets diluted, we have to manage, if we are the financial agent, we have to manage all that and the reimbursements make it difficult." One coded text on organization structure and capacity offered one individual's perspective on the challenges with the community-level intervention approach in a large city. There was a mixture. of suburban and urban organizations funded in each of the pools but this was the only mention of the challenges due to the size of the community. While this was in fact a response based on a misperception that those funded to implement the community-level intervention were mostly suburban, there may be merit to some of the issues raised and regardless, raises questions that may warrant further inquiry in a future study. This individual stated a belief that the community-level intervention worked well in smaller, homogeneous communities.
th en asked what the issue was m terms of doing the community level When ntl ·on this individual stated: interve The scope and diversity. Let me give you an example. Lets say you're working in one of the suburban communities and you can do a comprehensive approach for, I don't know, sixty thousand dollars annually. Well to do a comprehensive approach here is going to require half a million dollars." <Interviewer: "Because of different populations?> "And size of populations." <Interviewer: "So we are talking about basically it's the resources because of the diversity of the population?"> "Sure and our ability too. We have many more people to bring together to the table. If you want to go to a suburban community and you want a representative from the police department, you'll probably get the police chief and you'll get the superintendent of schools, and you'll get, you know, these will be all the decision makers at the It was also mentioned that the funding provided an opportunity to establish key relationships to solidify the coalitional structure and have the capacity to implement the intervention as well as enhance existing relationships with partners, other coalitions and the media. "I think we had developed some capacity at that point that we felt we would be able to get the necessary community entities on board to make it happen ... " One respondent said, " ... we also already had established a relationship with one of the police departments, and that really got enhanced greatly and we were able to reach out across community lines and develop a coalition and work with other towns." Another respondent commented, ''No, I think the network somewhat existed because we have the community partners, but it's just finding the appropriate partner for this particular grant." While several participants m Group 2 mentioned having pre-existing relationships with schools and the school department, primarily the necessary "linkages with other partners in. a coalition group in our area" to do a community-level intervention did not exist. It was also mentioned that it is too difficult and time consuming to establish those relationships especially in an urban area.
Six respondents in total felt they had the staff capacity to implement the programming for which they were seeking funding. An example of how staff capacity and previous experience factored into matching the organization with the funding pool, comes from an interview in Group 2: Yea, I mean it was the specific program we selected was a perfect fit because it required the same staffing, credentialing, experience, training, and oversight supervision, everything was really something we had a long history and experience doing, it was just doing it in another setting.
Other comments included: " ... we had the expertise to actually do the work in place." And, " ... we already had somebody trained with the appropriate qualifications and that was very critical, ... that's critical to the success of this particular model." Respondents were asked to rate their readiness in terms of capacity and resources at the time of the SIG announcement to implement a community-level intervention approach on a scale of one to five with five most ready. The seven respondents from Group 1 felt that they were ready to deliver the community-level intervention (M = 3.57) while the seven respondents in Group 2 did not feel as ready (M = 2.14).
The size of these organizations was different but consistent with the service delivery structure one would expect. Six coalitions in Group 1 had zero to four fulltime equivalent staff positions . (M = 1.2 FTE). Their annual operating budget at that time ranged from $34,000 to $150,000 (M = $78,250). One agency was excluded because it was the lead agency for the coalition. It had 55 FTEs and a $3.5 million operating budget at the time of the SIG announcement. The seven organizations in Group 2 had between five and 300 FTE staff positions (M = 156). Annual operating budgets ranged from $3.2 million to $30 million (M = $13.6 million).

Characteristics of the innovation.
Rogers' (1995) Perceived Attributes were the third most mentioned characteristic in the decision-making process. This emphasis however, may partly be the result of the grant structure and its multiple requirements which necessitated a number of questions to explore the advantages and disadvantages of each community-Jevel intervention component. The four Perceived Attributes which were the focus of this analysis were: Relative Advantage, Complexity, Compatibility, and Observability.
Trialability was not an option for organizations that were either funded to provide services or not.
Each of the four attributes was represented in participants' comments. There was some difficulty in coding these attributes as comments that were considered Complexity, Compatibility or Observability could also be considered Relative Advantage. For this reason, the concepts will be presented but not within a specific attribute.
There were a number of concepts that emerged as advantages and disadvantages of the community-level intervention approach, evidence-based programs, and environmental strategies. The most notable was the advantage of the community-level intervention being broad-based and comprehensive, "can target broader population in multiple ways." This approach, it was said, increased the likelihood for community-wide change, "multi-pronged approach that was going to be much more effective then what, then implementing one individual program." One respondent from Group 2 qualified their comment about the advantage of communitywide change with, "if you are successful." This last statement touches on the next t regarding the difficulty of managmg coalitions and community-level conceP interventions.
Respondents from Group 2 discussed the "difficulty" in managing coalitions and community-level interventions. Group 2 respondents did value this approach as the "ideal," but did not have the structure to implement it. One respondent described the community-level intervention as too "unwieldy to organize and manage, and implement and sustain." One specific aspect of this difficulty is the need for expanded relationships with other entities in the community and how that makes for a more complex process; especially for decision-making. The comments about difficulty illustrated the disadvantages for organizations in Group 2 that influenced and/or justified not pursuing the community-level intervention.
Group 1 respondents were more positive about the community-level intervention discussing its benefits and the opportunities it provided. One benefit of the community-level intervention approach was that it allowed for more services in the community. Three respondents said this approach provided an opportunity to build capacity specifically related to working with community partners and developing those relationships. One respondent stated: We always just focused so much on school and now we had an opportunity to reach out into other organizations and other avenues that we never reached before, like we were able to support more of the enforcement and taking a look at policy and you know, working more with the families and being in the media. I mean that's just something we never had the advantage of doing before.

44
Both groups shared similar advantages and disadvantages of evidence-based gram s which they were both implementing. These interventions had demonstrated pro evidence to support their effectiveness on specific health issues. Eight respondents mentioned this as an advantage in considering the outcomes they wanted to achieve, and in facilitating buy-in by other partners and staff who may have already been implementing other interventions that did not possess a foundation of evidence to support its effectiveness.
The strength of evidence-based programs was also their weakness. Adopting evidence-based programs eases implementation because the organization did not need to "reinvent" programming, everything was developed and provided. One consequence of "prescribed" or "canned" interventions was that organizations implementing these interventions must do so with fidelity to achieve similar outcomes; something that was "difficult" for at least four organizations and in one case impacted service delivery. A second consequence of "prescribed" programs was the compatibility between the test community with the serviced community in Rhode Island. For a couple of organizations this issue of fit with the cultural groups in their community was a concern. Implementing with fidelity left little room to adapt the evidence-based programs so some organizations were careful in selecting their interventions. For the number of organizations that were concerned with interventioncommunity fit, there were an equal number of organizations that commented on how the evidence-based programs were a "good fit" with their organization's goals and it was " a very natural extension of the work that we do." One final disadvantage of implementing the requirements of the community-I · tervention approach was the associated costs. While one respondent }eve in commented that these approaches were the best use of limited resources, four nd ents mentioned that it was expensive, particularly the evidence-based respo programs. One respondent shared the following about evidence-based programs: "The amount of dollars that it would cost to purchase the materials and carry out the program, also the cost of just the materials we found prohibitive and incredibly expensive." Another individual commented: .. .it was very costly, time consuming, and there was a lot of up front work that you had to do initially.... Once you get funded and then try to put all those pieces together, particularly multiple science-based programs, it's extremely costly to do.
Another comment about the financial disincentive was: The cost, the initial cost to implement it, to get our staff trained, to get staff familiar with the fidelity of record keeping, how you implement the program, the supervision for that. . I looked at a number of science-based programs and not chosen to pick because of the initial cost.

Communication behavior.
Communication behavior emerged as a major theme in the decision-making process. Communication behavior is a process of information gathering, sharing and processing "to reach a mutual understanding" (Rogers 2003, p. 5) which in this situation was deciding which pool of funding to pursue and which specific interventions to select for delivery. Information was sought through three channels: nal and mass media channels as well as through available scientific and local jnterperso data which emerged as a theme.

Interpersonal communication consisted of meetings and conversations internal
to the organization, meetings and conversations with peers and colleagues outside the organization, with program developers, and by opinion leaders.
Every respondent mentioned having conversations internal to the organization to discuss the different funding options. These organizations came to make their decision by having, for example, "a discussion about the two different options that were presented in the grant, and considered which we would be willing and able to pursue. " Meeting with colleagues and peers outside the organization contributed to the process of information gathering, sharing and processing. All but one of the respondents mentioned meeting with others external to the organization. The meetings and discussion were with other community partners such as schools, police, churches; other coalitions; state officials and meetings; and evaluators. The purpose of these conversations were for . additional information and input, to find out what other coalitions were planning to do, and sometimes to negotiate and get buy-in which was related to the opinion leader. Some of the coalitions that spoke with one another from Group 1 were in fact mentioning the conversations they had with one another.
Seven respondents contacted the program developers for information about the model programs, trainings, and cost. Three respondents had pre-existing relationships with developers.

47
Seven respondents stated that there were one or more individuals that were key in influencing decision-makers opinions. In five instances these opinion leaders were in fact the respondent. The opinion leaders were attempting to get buy-in from other partners such as schools and police for the community-level intervention. In at least one instance the opinion leader needed to overcome unfavorable attitudes toward evidence-based practices within their own organization. When asked how evidencebased programs would fit with the organization's values and norms, one respondent illustrated the role of the opinion leader when they said: It was a tough slog I got to tell you. I was really a lone voice on that and ... Twelve respondents used local data in their decision-making process including: RI SALT data, police and justice data, local needs assessment data, census data, and interpersonal sources for data such as coalition and community members as well as other organizational partners.

Prior conditions.
Each of the prior condition characteristics (e.g., previous experience, felt need, norms of the social system, time constraints) was mentioned by at least half the respondents.
Group 1 respondents were varied in their level of previous experience with community-level interventions and evidence-based programs.
For Group 2 respondents, several mentioned already having experienced delivery of evidencebased programs that the alternate funding pool would allow them to build on existing or previous work, "natural extension of our services to a different population." Several mentioned previous experience being part of or involved with a coalition.
There were a variety of needs mentioned across both respondent Groups. A couple of respondents stated they needed the money to replace lost funding. Others

49
. terested in providing new or expanded services. Respondents were directly were Ill d q uestion about the needs or problems that existed at the time of the SIG aske a announcement. The issue of need was rarely mentioned without this direct prompt.
This does not include comments about need in the context of agenda setting.
Time constraint was a theme 13 that emerged during coding and was not part of the original codebook. The SIG was a Request For Proposals which meant there was an assigned deadline. Some of the requirements for the community-level intervention funding pool were perceived to be too challenging to address within the time frame.
Three respondents in Group 2 mentioned there was not enough time to pull together a coalition; to do so within the timeframe was "impossible." For the community-level intervention funding there was a requirement that coalitions be in existence for one year, therefore a coalition could not be assembled. Community-based organizations could, however, have aligned with a coalition as a partner or lead agency. Two respondents from Group 1 also mentioned having difficulty with the timeline in preparing their application. They mentioned having only six to eight weeks to respond to the RFP which happened to . fall within the Thanksgiving -New Year holiday season.

Agenda setting.
Agenda setting comments were made by eight of the fourteen respondents at least once. Four respondents and their organizations said they had already identified areas of need and in some cases the intervention that would be the best fit prior to the SIG funding. The RFP provided the organizations with an opportunity to do something about their need. Rogers (1995) mentions "opportunistic surveillance" (p.
(Jr une constraint was added to the codebook as a subtheme for the "prior condition" theme.

50
) an activity where organizations search for innovations or opportunities and 393 as try to match them to a problem/need rather than searching for a specific solution then rob lem This description may align more with some coalitions and communityto a P · nnizations and the nature of their funding source. orgaiu Three respondent's organizations didn't identify an area of need as much as react to a loss of funding. The loss of funding and gap in services influenced the decision to apply for funding and in one case, may have pre-empted the organization and community from taking time to identify its most important needs: you know there is always a squeeze because sometimes organizations, and I know we got caught in this and I think the community at that time got caught in this, when funding for one program dies out which I believe at that time we lost a position in [school name] because the funding stream ended. So we were thinking about replacing that and that probably constrained our thinking about what we should do, if not constrained that might be too strong a word, but it influenced our thinking about what we should do realizing that there would be a hole in our prevention. network at [school name] if we did not do anything to replace that position. So I think we were driven more by those concerns than we were by, what does the community really need at this time, what could, should we be more comprehensive.
Knowledge. Rogers (1995) describes three types of knowledge (i.e., awareness, how-to, Principles). There was some mention of awareness knowledge but it was typically awareness about the availability of funding. Information about the innovation was h through a previous grant, training or work experience and while not exactly approac d as a community-level intervention these experiences were somehow similar stfUCture gh that respondents drew a connection between these experiences and the enou conununity-level intervention as it was designed in SIG.
While the types of knowledge described by Rogers were not highly stated by respondents, the information gathered through communication channels certainly provided "how-to" and possibly "principles knowledge;" although the participants were not asked what type of information they were seeking.

Decision.
There was a difference between Group 1 and Group 2 in the innovationdecision type they used which can be explained by their structure. Group 1 primarily used a collective decision type while Group 2 used the authority type and contingent type (i.e., combination of authority and collective).
Two coalitions in Group 1 used the collective type exclusively while an additional two used the authority type after having made a collective decision. In both instances it was in response to. the timing of the RFP and the short timeline for submitting an application. Convening or contacting the coalition membership to make quick decisions in the short timeframe would have slowed the grant writing process.
Three respondents in Group 1 described making an authority type decision.
Two of these organizations were lead agencies for the funded coalitions and two had the Mayor as the final decision-maker.
Four respondents in Group 2 used the authority decision type and three used the contingent type (i.e., collective and authority). For all seven respondents, the final . . rested with one individual in an executive management position and may dec1s1on 1iave even gone to a Board. The decision process for these respondents involved making recommendations to a top executive. The difference between the two decision types is based on who was making the recommendations: other individuals in management or program staff and/or committees.
Coalitions with few if any paid employees and a volunteer membership that essentially serves as a community board is structured in more of a non-hierarchical manner so that making decisions by voting or through consensus fits. They may meet once a month to discuss coalition issues and make any necessary decisions.
Community-based organizations are typically hierarchically structured which lends itself more to decision-making by executives.
Two general examples of the decision-making process for community-based organizations are presented to briefly highlight that the process is multi-layered and is much more than just one or a few executives saying yes or no. One respondent described the process of attending the Community Readiness Training and becoming aware of SIG funding. This individual then went back to the organization and met with the Program Director to discuss the possibilities and was given a go-ahead to begin doing research. The respondent and Program Director met again to discuss the research and made the decision. In another example the respondent went to the Executive Director, received a go-ahead, did the research, disseminated the research to all relevant staff who then met internally to discuss the options. The group decided to Pursue the single evidence-based program. The group then discussed which specific evidence-based programs would be best. Meanwhile the respondent went to the 54 1 t meet with the school principal. The principal disseminated the information scho<> 0 to school personnel who then met internally to discuss the proposed funding arrangement. Tue school personnel and respondent met together to discuss the funding and decided to work together.
Coalitions in Group 1 were faced with similar circumstances needing to identify and select appropriate community-based organizations to make arrangements for the delivery of the evidence-based programs. The environmental strategies were a similar challenge in trying to engage other community partners that may not have been actively involved with the local coalition.
Six Group 2 respondents said that the same decision-making process was used in deciding which funding pool to pursue. These organizations considered both options but for most, the community-level intervention funding pool was an easy and quick decision to not apply. The major influence was the coalition structure requirement. One respondent described it as not having "the resources to organize a coalition and do the broader community level approach." The frrst funding option was too "difficult" but the seconcj option, the single evidence-based program, was "manageable" and "more realistic." For the Group 1 respondents, they mentioned that they had the coalition structure which one respondent described as a "very strong influence" in pursuing this funding. Six of the respondents stated that they wanted to do comprehensive services, that with these services they could reach the community on a broader level. "It was a great mechanism because we really looked at it as [a] multi-pronged approach that was b much more effective than what, than implementing one individual going to e program· " A couple of Group 1 respondents wanted to do comprehensive services even though they did not know much about environmental strategies, That requirement, as I recall, wasn't well understood at that time. What the environment was. I mean no, the environmental strategy requirement. ... no one knew what that, what the environmental program really entailed . ...... so I don't think anybody really knew much about it. .. .I think we knew that we would just be exposing ourselves to a new sector.

Research Question 2
Tue second research question was to determine which characteristics were most influential in the decision to adopt or reject the community-level intervention.
The analysis was conducted in two parts. The first analysis reviewed the participant's responses to one interview question. The second part answered the question of most important characteristics by presenting the characteristics that were mentioned most often by respondents.

Non-guided Analysis for Research Question 2
Participants were asked which factors were most important in influencing their organizations' decision to apply or not for the community-level intervention funding.
This interview question was designed to ask about the decision process in an openended manner, without providing any particular prompt for the respondent, thus allowing the respondent to offer the underlying reasons for their decisions with minimal bias introduced by the interviewer and interview process.

56
The most important characteristics were matching m association with cbafacteristics of the organization and relative advantage. Organization complexity was an important characteristic for Group 2 participants in applying for the other pool of funding (see Table 5).
Group 2 (Interview #) 1 5 6 8 14 15 16 0pinion leader x Contact with developer x Matching the requirements of the community-level intervention funding with the organizational characteristics, target audience, and identified needs were important factors influencing the decision-making process. Matching was overtly stated by only two participants (e.g., "fit" or "match") but it is the implied frame within which the organizational characteristics, target audience and identified needs are stated.
The level of cooperation and collaboration 14 was mentioned by three participants. Two comments from Group 1 participants were, " .. .I think the other thing that prompted them, I think, was the high level of cooperation and collaboration that had previously occurred, so they felt that it could work." And, I think the network somewhat existed because we have the community partners, but it's finding the appropriate partner for this particular grant.
<Interviewer: So the grant provided you an opportunity to do that?> Yes.
<Interviewer: To find the appropriate partners?> Yep and to collaborate ....
The single respondent from Group 2 mentioned that they did not go for the community-level intervention funding because they, " ... just didn't have those linkages with other partners in a coalition group in our area." Organizational structure and capacity was mentioned by five respondents (two from Group 1 and three from Group 2). Analyzing the responses across the Groups 14Th.
did not reveal any concepts. There was a concept among Group 2 respondents which ro ss coded texts from organizational structure and capacity, relative advantage, cut-ac and the innovation complexity themes. There was an underlying reason for the majority of Group 2 respondents in not applying for the community-level intervention funding; that they were structurally not coalitions. The following coded text is an example of the perceived difficulty in becoming or being involved with a coalition in order to have applied for the community-level intervention funding, "Just the difficulty in forming those coalitions and then the funding gets diluted, we have to manage, if we are the financial agent, we have to manage all that and the reimbursements make it difficult." Three respondents (one from Group 1 and two from Group 2) mentioned alignment of the funding with the population they had identified and two respondents from Group 1 associated the funding with the pre-identified needs for their target audience (agenda setting).
Organizational complexity was mentioned by three participants from Group 2.
The common concept among these individuals was having trained or qualified staff for the programming that was available in the non-community-level intervention funding pool. These organizations matched their staffs previous training and credentials with a specific evidence-based program.
Relative advantage also emerged as an important factor in organization's decision-making. One of its subdimensions is providing a cost/economic benefit.
Money or funding was mentioned by three participants in Group 1 and two Participants in Group 2. For the Group 1 respondents the funding benefit was 59 . t d with an opportunity to provide more services, "The fact that we would be assoc1a e I t d o more programming and be funded to do that." Another respondent stated, abe o "Well the opportunity to use funding to address the issue of substance abuse which the group had already identified as a significant problem." This was also stated by one of the Group 2 respondents who commented, " ... to have a funded program that required parents to participate." A different respondent from Group 1 made a comment that was not directly associated with funding but which further reinforced the previous concept, " ... having an opportunity to implement something. To implement an evidence-based program." These comments indicate that participants saw an opportunity for additional programming made possible by the SIG.
There were two other comments made related to funding; both by Group 2 respondents. Their comments intimated a desire to maximize the grant award for their organization as a reason for not applying for the community-level intervention but for the other funding pool. For example, one individual said, " ... and then the funding gets diluted ... " while the other stated, "That I could largely keep it in house."

Guided Analysis for Research Question 2
An alternate approach for considering the most influential characteristics was to identify the characteristics most often mentioned by respondents. The more frequently a characteristic was mentioned, the more important that characteristic was to the decision-making process. To determine the most frequently mentioned characteristics, the findings presented in Table 4 for the guided analysis for research question 1 was used.

60
The most frequently mentioned characteristics included matching the benefits and disadvantages of the innovation with the organization, community and needs; .... ;zation structure and capacity (e.g., being a coalition or not); organization system orgcuu openness (e.g., having interorganizational linkages or not); organization complexity (particularly for Group 2); communication through interpersonal and mass media channels, using available scientific and local data; and the relative advantages and complexity of the model (e.g., broad-based and comprehensive approach reaching entire community, too difficult to manage especially when structurally not a coalition).

Research Question 3
The third research question was to determine whether the characteristics that influence a decision to adopt differ for organizations representing culturally diverse communities. Despite an attempt to oversample organizations serving culturally diverse communities, only four organizations were interviewed.
One of the organizations had applied for the community-level intervention funding pool. Two applied for the alternate, single evidence-based program funding pool. The final organization applied for another alternate funding pool to implement a locallydeveloped intervention. The transcript from this final organization had not been used in the analysis for the previous two research questions due to a language barrier in conducting the interview.
A content analysis was conducted on the transcripts of the four organizations that were identified as serving culturally diverse populations. The analysis reviewed Participant's responses to one interview question that did not guide them through the characteristics. Responses to this question would highlight the most important factors influencing the decision to apply or not for the community-level intervention funding similar to the second research question. This analysis using four transcripts and one interview question did not provide enough coded texts to identify consistent themes.
A follow-up analysis was conducted expanding on the number of questions from the first attempt. A matrix intersection Boolean search was conducted to crosstabulate text coded as characteristics of the organization and characteristics of the innovation pairwise by questions 2.9a through 3.14. These questions were selected because they ask about the decision process in a less-directed manner; allowing the respondent to offer the underlying reasons for their decisions with minimal bias introduced by the interviewer and interview process.
Again, with so few organizations interviewed it was difficult to identify themes within the group and any differences in comparison to other organizations.
Preliminary findings indicated that the characteristics cited by the organizations serving culturally diverse populations were no different than the other organizations (see Table 6).
There was one possible theme among the four organizations. Compatibility, a characteristic of the innovation, was mentioned three times by two respondents. Their responses were specific to the issue of using an evidence-based program developed elsewhere with culturally diverse populations in Rhode Island. The coded texts are not quoted here to protect the confidentiality of the respondents.  Note. Cod d t . e exts are only from questions 2.9a through 3.14 of the survey.

Research Question 1
This dissertation sought to determine if Rogers' Innovation-Decision Process 01 odels "fit" the self-reported decision making processes of coalitions considering whether to undertake an innovative form of prevention programming. From a "pattern 1Datching" perspective, the answer is a resounding "yes". Every stage from Rogers' individual and organizational models that were studied were mentioned by the respondents. The majority of characteristics associated with these stages were also 01 entioned by the respondents. Furthermore, constructs outside of Rogers' models that have been proposed by other researchers were, with one minor exception, not mentioned. This study can thus be seen as a further validation of the "nomothetic net" (or conceptual framework) established by Rogers and adds to the evidence presented by Rogers (Rogers, 1995) that his models have broad applicability to a number of content areas including agriculture, business and industry, healthcare, education, and local health departments.
The most salient constructs m the decision-making process for the organizations in this study were matching in association with organizational characteristics (structure, capacity, resources, complexity, openness, goals and mission) and organization and community needs; previous experience; relative advantage; and time constraints. Using the coded texts from the two primary, nonguided questions, the matching stage emerged as the major point of association for Various characteristics.
Respondents from Group 1 expressed a belief that they had the capacity to effectively implement the community-level intervention and saw it as an opportunity to offer additional services. Part of this capacity was the relationships with key cororounity individuals and sectors that the coalitions had established. Furthermore, the characteristics most cited as matching by the organizations in Group 1 coincided with the first three of Florin, Mitchell and Stevenson's (1993) coalitional developmental tasks (i.e., initial mobilization, establishing structure and building capacity for action).
That is, many respondents mentioned having multiple community sectors represented on their coalition, or referred to being an established structure with a decision-making procedure, had staff expertise and community linkages to get things done. Some of the coalitions also saw adopting the communitylevel intervention as an opportunity to broaden sector representation, develop new relationships, strengthen existing relationships, and in general strengthen the overall coalition structure and capacity. In a symmetrical way then, coalitions can build community-level interventions and community-level interventions can build coalitions.
Conversely, the organizations in Group 2 clearly commented that they did not have the structure or links to coalitions to pursue funding for the community-level intervention. Applying for the community-level intervention would have presented a disadvantage for Group 2 organizations, in that they would have had to alter tlwjr ' structure or relationships to fit the funding opportunity. This was something they obviously did not need to do to apply for the alternate funding pool.
Group 2 organizations were larger in size and had more financial resources tba.Il their counterparts. Usually one would expect the larger organizations to be more inJlovative since they have more capacity in general.
In reality the larger organizations opted to apply for the less challenging and lower dollar amount funding pool. In the context of this study, matching the organizational structure (i.e., community organization vs. coalition) to the specific funding pool was the most salient factor.

Research Question 2
This study also asked which specific stages and characteristics were most important in the decision making process. Using the number of coded texts as a general indicator, the most important constructs, in order, were matching, characteristics of the decision-maker, characteristics of the innovation, communication behavior (overlapping the knowledge & persuasion stages), prior conditions, and agenda setting.
Asked in one open-ended, non-guiding question, important characteristics were matching in association with characteristics of the decisionmaker, target audience, and identified needs (agenda setting); relative advantage as a cost/economic benefit; and organization complexity (for Group 2 respondents only).
Decision makers from both groups closely considered the "match" between their existing organization and the demands that would be imposed by adopting the innovation. Matching was associated with the characteristics of the decision-maker, particularly organization structure as a coalition or not and level of collaboration and cooperation with other organizations in the community (organization system openness).

66
The importance attached to relative advantage, a characteristic of the tl ·on in this study mirrors Keam's (1992) finding mentioned in the introduction inflOVa ' that such variables were far more powerful than a host of other characteristics in predicting adoption of an intervention. Relative advantage was specifically mentioned in the context of the funding opportunity providing a cost/economic benefit.
Particularly for Group 1 respondents this opportunity allowed them to provide more services within their communities.
For Group 2 respondents, having trained and qualified staff (organizational complexity) influenced their decision to apply for the alternate funding.

Research Question 3
One of the major benefits of implementing an evidence-based program is that higher implementation fidelity increases the likelihood that positive outcomes similar to those achieved by the developer can be attained. In fact the demonstrated effectiveness (observability) of evidence-based programs was an important factor for some who saw the "packaged" programs as an incentive that reduced curriculumdevelopment work. The issue . is that evidence-based programs are sometimes developed and evaluated with populations that are different from those that will participate in the program after its dissemination; an incompatibility. This places community-based organizations and coalitions in a position of making an adoption or rejection decision using incomplete information. Indeed, the shift towards the coalition, community-level intervention approach may come at some cost for organizations serving culturally diverse populations, or at the very least, a difficult decision to be made. This was certainly evidenced by the responses of one · servicing culturally-diverse populations. The saliency of innovation organization coropatibility for organizations serving culturally diverse populations was difficult to ascertain with so few coded texts. The incompatibility of evidence-based programs that have been tested elsewhere with different population groups was mentioned by two organizations.
Additional Findings: Stage Models Rogers (1995) mentions that the concepts from the individual process contribute to the organizational innovation-decision process, but, "when an innovation-decision is made by a system, rather than by an individual, the decision process is more complicated because a number of individuals are involved" (p. 22).
This was consistent with the general findings of the guided analyses that the early stages in both the individual and organization innovation-decision process models were supported within the organization decision context, thus adding to the body of evidence regarding the existence of the different stages.
Beal and Rogers (1960) provided empirical support for the existence of stages and that there was a progression through each of the first three stages of the individual innovation-decision process. There were other studies cited with different occupations or fields supporting the stage sequence. However, Rogers (1995) did state, "The evidence is most clear-cut for the knowledge and decision stages and somewhat less so for the persuasion stage" (p. 188). At that time he also stated that there was "rather poor data" in support of the later individual stages (p. 188).
Upon closer inspection using the non-guided analyses, it may be the case that the decision-making process for organizations was in fact a communication process with matching as the central activity; that the individual knowledge and persuasion stages as well as organization agenda setting stage were embedded in the central .. n;zational activity of matching. Taken together, the results from this study, at org ..... least in this context, suggests less a progress through a sequence of steps and more a simultaneous, organic process that simultaneously considers the fit of prior conditions, characteristics of the organization and the innovation, and needs in making a decision.
A conceptual model intended to reflect the matching process of coalitions and community-based organizations in the context of this study is presented (see Figure 3).
A successful match would lead to a decision to adopt and implement. What is not reflected is how other organizational partners (e.g., schools, other community organizations) may or may not be engaged in the adoption decision process. This includes their internal decision process in matching with whatever scenario is presented to them in regards to the adoption and implementation of the innovation.
Influential factors from other external systems are also not portrayed (e.g., time constraints).  Rogers (1995) describes the decision-making process as a communication process that involves information gathering, sharing and processing in the formation of an opinion about the innovation among members of a social system towards a common understanding and decision. Rogers also suggests that mass media channels are "relatively more important at the knowledge stage" while interpersonal channels are more important for the persuasion stage (p. 195). The challenge was that some of the individuals interviewed spoke with their peers to both increase their knowledge and sh&pe their opinion. Gathering and processing information also seemed to occur on an ongoing basis as needed. On some occasions the individuals who were tasked with gathering information would hold meetings to share what they had found.
Learning new information without processing whether it reinforces or is incongruent with a held attitude or belief would seem to be difficult. The act of searching for new informatioh about an innovation may even be through a pre-framed lens of the '70 individual's attitudes and beliefs.
In any case, the communication behavior un tered in this study seemed a continuous cycle at times that made it difficult to en co separate and code text into the unique knowledge or persuasion stages. This excludes the communication channels which were distinct and easy to identify although not in context with the knowledge or persuasion stages. It also excludes the characteristics of the decision-maker and innovation which, as mentioned previously, were strongly associated with the matching stage.
Communication channels were mentioned in the introduction as an important characteristic since most individuals form their opinion based on peers who have adopted rather than "on the basis of scientific research by experts" (Rogers, 1995, p. 36). Respondents in this context, however, used scientific and local data in gathering knowledge about the innovation to make a decision and for planning. This may be the nature of the field and specific to the context of this study. Nevertheless, it does challenge Rogers' statement and would seem to indicate the general adoption of an idea of evidence-based practice or decision-making (i.e., these community organizations and coalitions approach their work taking into consideration the scientific evidence-base for their practices or programs that are being adopted).
A study by Meyer andGoes (1988, as cited in Rogers, 2003) found that the characteristics of the innovations (i.e., perceived attributes) explained 40% of the variance while characteristics of the adopting organization only explained 11 %. This and other studies were interested in how the organizational characteristics predicted organization innovativeness.
This study was not concerned with predicting organizational innovativeness and fit more with what Rogers (1995) described as SS re search" (p.188) instead of the "variance research" (p.188). This study did "proce d . te from Rogers' definition of process research in that it was not explicitly ev1a investigating the "time-ordered sequence" of the stages (p. 188) although there was evidence to suggest that the process was more a simultaneous, organic one in the It may also be worth redefining relative advantage and the other characteristics of the innovation. There were some instances where coding text was difficult. The common denominator was that respondent opinions were all advantages or disadvantages of the innovation. . For example, respondents spoke of the benefits of evidence-based programs as having demonstrated effectiveness. This fits with the observability attribute but it is also a relative advantage because this demonstrated effectiveness is perceived to be better than what is currently being implemented. It has been stated by others that these attributes are intercorrelated but Rogers holds that they are not.
The communication and decision process was further complicated by the nature of the social system within the community. The coalitions adopting the 72 I conununity-level intervention needed to fit the innovation not only with their needs and organizational structure but also with the needs of the community members as well as other community organizations and institutions (e.g., schools, police). The community coalitions needed to match with community partners they felt had the capacity to deliver the specific interventions. Some organizations worked around partners that they believed would not easily participate. The coalition decision process was actually multi-layered and could be said to require a second tier of the innovationdecision process with the organization (or its individual representative thereof) as the opinion leader in actively lobbying for the outside organization or institution to participate. This additional layer in the decision-making process is unique to this specific context and places a systemic constraint on the organization from an external source. These other organizations or institutions will have their own set of characteristics and needs to consider regarding the innovation as well as regarding the primary adopting organization that they would have to partner with.

Additional Findings: Systemic Barriers
Felt need and agenda setting were combined during the analysis of coded texts because they were conceptually similar. While some organizations mentioned having identified their needs prior to the announcement of SIG funds, others may have taken an "opportunistic surveillance" approach. Rogers (1995) described this approach as organizations engaging in "scanning the environment for new ideas that might be beneficial to the organization" (p. 393). Felt need and agenda setting or opportunistic surveillance was a challenge for this study because these organizations reacted to a funding opportunity to provide community services. The RFP scope of work and l ·cation window were dictated by an outside entity. The mismatch is that the app l funding itself is not the new idea; the community-level intervention was the innovation. The funding was for a community-level intervention but it could have easily not been provided as an opportunity. The community-level intervention as an idea or approach still exists but had it not be offered with funding, chances are that the coalitions would not have implemented this approach. Without financial support, it is unclear whether the coalitions will continue to implement this approach.
The organizations in this study are dependent on grant funding which introduced a new characteristic of a time limit. This characteristic is not related to the advantages or disadvantages of the innovation nor is it really a function of the characteristics of the decision-maker. For example, the time constraint has nothing to do with whether the organization has the appropriate structure and capacity to implement a community-level intervention. It could be said that this characteristic is related to organizational structure and capacity in that the organization must have the organizational capacity to apply for funding, but this is not an issue directly stemming from the innovation itself. The community-level intervention approach can be adopted by an organization as their approach of choice, but unless the organization is creative in translating a collection of different funding streams into this approach "on the ground," the organization is still dependent on this outside entity to even offer an opportunity for funding to implement this approach. It is not only a time constraint in the window that these organizations are given to apply, it is also indicative of a systemic constraint that non-profit community coalitions operate within.

Implications of Findings
The most notable difference observed in this study was in the decision-making strle between coalitions and community-based organizations. Coalitions were more likely to use the collective decision-type approach while community-based organizations were more likely to use the authority or contingent decision-type approach. There is no value judgment to be placed on which is better. Each organization is structured differently and the type of decision approach fits with that structure. These different decision types correspond with different flows of internal communication and information. The coalition coordinator usually presented the funding opportunity at a meeting of its members and they discussed it and either reached a consensus or voted. In two instances coalitions had to shift to an authority type after the initial collective type decision due to logistical factors with the RFP timing. The community-based organizations, as one example, had more layers of meetings, approvals, and recommendations before an executive made the final decision. Due to the nature of the interventions, both groups needed to engage other community partners. The difference was in the number of partners that needed to be engaged.
There was an internal and external communication process that was multilayered. Rogers described it as the contingent decision type; that combinations of decision types might take place or that one decision can only be made once another has made their decision. Because these organizations serve community populations they must work with or have arrangements with other partners in the community to deliver services. For this reason the decision process happens in multiple layers. Some organizations described needing to get the schools, for example, on board before the decision could be made to go for the funding. In at least one instance the communication-decision process occurred in the school setting as well as in the organization. For some coalitions, they may have needed to negotiate with multiple partners who may have had to decide in their organizations whether aligning with the coalition would be beneficial. The relevant issue is that the communication-decision process may occur among a number of partners in this funding and organizational structure, each contingent on the other.
The applied implication for state agencies is the need for built-in training and intervention piloting at the front end of contract awards. Considering that time constraints emerged as an issue for several respondents, providing sufficient time and support for organizations to meet the requirements of challenging RFPs would also be advised. The funding opportunity according to some respondents did provide an opportunity for capacity building but it was not built-in. The Strategic Prevention Framework (SPF) SIG was mentioned as an example of the type of predecessor funding that would have assisted coalitions in applying for and implementing community-level interventions.
A critical sentiment shared by some of the respondents was a sense that with the award of the SPF SIG the original SIG was less supported.
For Group 2 organizations, the community-level intervention was very specific in its requirements and while respondents spoke to issues of difficulty and it being a disincentive to link with coalitions, some did see the value in the approach. Still others mentioned seeing the value in the approach now, years later. If the goal for the state funders is to have community-level interventions within all municipalities then taking time to help foster and support relationships between community-based organizations and coalitions may be warranted.
For two respondents the use of an intermediary was helpful. A technical assistance structure using intermediaries to support coalitions prior to applying and then throughout the process of strengthening their coalition and implementing the community-level intervention may be a promising approach worth further exploration.
The implication of the actual decision process and whether it makes a difference in practice is unknown. The organization's internal process is structured in a manner that works for them.
In summary it is suggested that community coalitions, community-based organizations and the state need to do a better job of creating truly integrated community-level interventions for implementation. In each Rhode Island community that has a coalition, that coalition should be working with or have inter-organizational links with other community-based organizations providing social services. The findings of this study would indicate that in some communities this link does not exist or could be strengthened. Granted some organizations were linked or had been but were not interested in pursuing or partnering for the community-level intervention. To build community capacity it will be necessary for all community-based service providers to communicate in the development of community needs and assets assessment, and in the development of an integrated workplans for addressing community needs. The state can encourage these activities by providing training and technical assistance to coalitions to help them serve in the role of community 77 I I convener, how to reach out to other organizations, provide the skills necessary to conduct the needs and assets assessment and complete a community-wide workplan.
The state is also in a position to adjust their RFPs by incorporating planning and implementation phases as a standard practice. This would maximize efficiency by providing time for communities to conduct the aforementioned activities. Those that demonstrate success would then be eligible for funding in the implementation phase.
Ongoing skill-building support for those in the implementation phase would still be necessary. Multiple state agencies could also form a partnership to establish a more integrated service delivery model at the community level.

Limitations
There were several limitations to this study including issues with recall, methodology, and other biases. Rogers (2003) suggests collecting data at multiple time points before, during and after the individual or organization makes a decision. A major limitation of this study was the use of retrospective, self-report recall data at one time point (Rogers, 2003). Depending on the length of time that has passed respondents may have errors in recall, accuracy, and may frame their responses by the success or difficulty they have experienced in implementing the innovation (Rogers, 2003).
Respondents were asked to recall the SIG communication and decision-making process which had occurred approximately five years earlier. Most respondents clearly remembered the process and as some were interviewed they began to remember more. There were certainly a number of individuals that absolutely could not remember including those that were not funded through SIG. At some point during the interview, most respondents commented that the SIG process had been years ago either overtly stating or intimating their struggles with recalling information to answer the questions. Rogers (1995) states that the decision-making process is a social process that occurs over time and distinguishes it from other decision-making processes in that the decision is about something new and that there is an uncertainty involved in adopting an innovation as a "new alternative to those [other innovations] previously in existence" (Rogers, 1995, p. 161). The majority of respondents in this study said they were aware of this approach prior to the funding announcement. The implication is that this was not a new innovation. Most, however, had not used this approach in the manner it was structured within the SIG. This was further compounded by the fact that the coalitions and community-based organizations had a choice between the community-level intervention and the alternate funding pool, which for many in Group 2 was an extension of work they had already been doing. With the choice of two funding opportunities, these organizations were not faced with applying for one or nothing at all. If in fact only the community-level intervention funding pool had been available it is unknown how the community-based organizations in Group 2 might have acted. Their need for funding may have forced them to opt for the communitylevel intervention. Ultimately, there was a level of uncertainty that was removed by having the multiple funding options.
The main methodological limitation was the lack of an intercoder reliability check on the codebook and coded texts. Ryan and Bernard (2000) state: "The coding of texts is usually assigned to multiple coders so that the researcher can see whether 79 I ' I the constructs being investigated are shared and whether multiple coders can reliably apply the same codes" (p. 785). While the codebook was developed from Rogers' pre-defined and researched characteristics, no reliability check was used in the application of the codebook thus limiting the external validity of the findings.
Other methodological limitations included using the comments of one individual to represent and generalize to the entire organization when many more may have been involved in the decision (Rogers, 2003). Another was not having more experimental groups that were aware of but did not apply for either funding option.
These organizations may have provided a different perspective on the decision-making process. An attempt was made to interview some organizations that fit this description, however, not having gone through the application process or not having implemented the interventions, in combination with the amount of time that had passed; they did not remember the SIG.
Some limitations with measurement and analyses were also encountered. For example, the characteristic of organization structure and capacity can encompass other characteristics of the organization. This characteristic and the others were separated and defined but it can still present some challenges when interpreting the results.
The survey was constructed to elicit responses based on Rogers' model. The questions may have influenced the responses and given the perception that Rogers' model was applicable when in fact it was an artifact of the questioning. An attempt was made to protect against this by beginning with non-guiding questions.
The codebook was developed from Rogers' stages and char&cteristics and applied to the transcripts. A grounded theory approach could have been used to review the transcripts and allow the themes to emerge and then compare it to Rogers' models.
This was primarily a qualitative survey and analysis. In developing the measure, coding the texts and conducting the analyses, the biases of the author would be present. One concrete example of how the author's bias may have been introduced was in interpreting the results related to the matching stage. There were instances when it was not clearly or overtly stated that organizations were matching. Sometimes the matching was intimated but it was the frame within which the organizations assessed their issues of fit in relation to the funding requirements. Also, some of the respondents repeated themselves across questions during the interview. The author tried to code unique texts but some of the coded texts may in fact reflect repeated statements across different questions. This was one reason why it was mentioned that the numbers of coded texts in the tables should not be compared on a similar scale, rather, the number of coded texts serve as indicators for further review in the pattern matching approach. The important point was whether respondents mentioned the same characteristics in the same context.
Some limitations of this study are related to diffusion theory and the decisionmaking process. One of the major criticisms of diffusion theory is what Rogers (2003) terms "pro-innovation bias" (p. 106). The premise of this bias is that the innovation is good, should be adopted as quickly as possible by many, and should be implemented with complete fidelity with no adaptation or reinvention. For example, some States used their SIG funds for the implementation of single evidence-based programs only.
In Rhode Island, potential users were given funding choices that included the aforementioned as well as the community-level intervention. Rogers (2003) in citing JD Eveland (1979) notes that a pro-innovation bias is not inherently wrong and the jnnovation may in fact be in the public's best interest. One method for overcoming this bias are to collect data at multiple time points which was not the procedure for this study (Rogers 2003). Another method is to ask "why" questions: "We should increase our understanding of motivations for adopting an innovation. Such "why" questions about adoption have seldom been probed effectively by diffusion researchers" (Rogers 2003, p.115). These types of "why" questions were used in the interview and analyzed separately for the first and second research question as non-guided responses. The results of these analyses were presented.
Rogers' (2003) models focus on decision-makers as mostly free to decide whether to adopt or reject an innovation on the merits of the innovation with minimal consideration for pressure to adopt exerted by external sources. In this study the state of Rhode Island provided funding to community-based organizations and coalitions for the implementation of a service that would benefit a local community. These services were designed to chang~ health behaviors but not for the organization that decided to implement the service. The individual service recipients were the decisionmakers who had to decide whether to adopt and implement the services. The coalitions in delivering the community-level intervention were somewhat of an innovation broker between different social systems. The different organizations in the social system were not interviewed.

Future Directions
Some potential future directions might include changes to RFP structure especially when introducing new approaches that are very challenging. Providing training, technical assistance and support prior to and throughout the process would be beneficial. Using intermediaries may also be a promising approach and warrants further inquiry into other states that may use a similar model. Building the coalition development activities at start-up into the first part or phase of a grant might be beneficial and would minimize issues expressed in this study as time constraints.
One possible future study might specifically focus on issues of adoption of the community-level intervention, evidence-based programs, and environmental strategies during the grant process and with some comparison groups. This would seem appropriate given the findings of this study. The value would be to identify the reasons why some coalitions may not pursue community-level interventions and the factors that influence a coalition's perceived ability to implement a community-level intervention.
One respondent commented that the community-level intervention was easier in smaller homogeneous communities vs. diverse urban areas. This respondent said it would be cheaper in the smaller areas and that in the smaller, more homogeneous areas they are able to get the appropriate partners to the One other study would be to investigate the implications of those organizations that identify their needs (i.e., agenda setting) versus those that use opportunistic surveillance and how this might relate to their decision-process, if at all.
One final future direction may be to expand the scope of this study to include the "second tier" of innovation-decision processes among partnering organizations and institutions. The need for community partners was a necessary component of the community-level intervention and one that some coalitions had been concerned with at the beginning of the grant award, so-much-so that one organization selected a specific type of evidence-based program to work around or remove an organization from the process. The decision-making process in the community reflects a more complex network of decision-makers and certainly warrants further inquiry. well as on tape for later transcription and entry into a statistical software program. The paper copy will be stored in a locked file cabinet in my personal office at the Maine Center for Public Health in Augusta, Maine. The Center's office has an ADT security system and the only entrance into the office is locked at all times, even during business hours. The cassette will also be stored in the locked file cabinet until it is transcribed at which point it will be destroyed. All electronic files are on a firewall and password protected network. The electronic interview information will be in my personal drive on my office computer which is secured with my personal password.
If at any point you have any questions or concerns you may contact me directly at 207-629-9272 x209. I am more than happy to discuss this study with you and address any issues you may have. and you may stop at any time. I do not believe there is any risk of harm to you in answering these questions. You will not be identified by name in any report; results will be written in a way that protects the identity of individuals. This interview will be tape recorded but the cassette will be destroyed as soon as it is transcribed. Should you have any questions or concerns now or in the future please contact Marco Andrade at 207-629-9272 x209. Thank you for agreeing to participate; it is my hope that sharing your experiences will inform future Characteristics of the decisionmaker -organization system openness Successful adoption and implementation of innovations on previous occasions will have a positive impact on future decisions to adopt innovations and on the success of future implementation Decision-makers must experience a felt need or problem to be open to messages about an innovation. This perceived need should be sufficient to mobilize an individual, organization or community to search for a new solution (Price, 2000). There are, however, occasions when knowledge of an innovation precedes the felt experience of a need. Org climate; buy-in The potential user must possess the capacity and resources required to successfully adopt and then implement an innovation. For an organization these might include: size, degree of formalization and centralization, trained staff, staff availability, and financial resources. "The degree to which an organization' s members possess a relatively high level of knowledge and expertise, usually measured by the members' range of occupational specialties and their degree of professionalism" (Rogers, 2003, p. 412). This is ex ressed by formal training. This is the best predictor of organizational innovativeness (Rogers, 2003). This characteristic encompasses the number of staff, size of budget, size of community they serve, prestige and social influence of the decision-maker among their peers, credentials, staff technical expertise, and cosmopoliteness as evidenced by such things as how many boards and committees one belongs to outside the organization. "The degree to which the members of a system are linked to other individuals who are external to the system" (Rogers, 2003, p. 408) Characteristics of the decisionmaker -readiness

Knowledge -awareness
Knowledge -how-to Knowledge -principles Communication behaviorinterpersonal channel (peers/outside orgs) Community sectors; developing partnerships The individual, organization or community believes that the identified solution will address the problem.
The individual, organization or community perceives its readiness for, and capacity to manage change that will result from the adoption of a new solution.
The individual, organization or community is willing to increase its readiness for change if necessary. Receptiveness/o enness to innovation When a potential decision-maker first becomes aware of an innovation before seeking information about that innovation. These decision-makers may learn of an innovation's existence from change agents or through colleagues (social networks). Others who experience a perceived need or problem may proactively search for a potential solution using whatever resources are available and accessible (e.g., "What is the innovation?") or knowing that an innovation exists (p. 165). Rogers (2003) states that, "at any given point in time, many potential adopters are aware of a new idea, but are not yet motivated to try it" (p. 213). Think about awareness of RFP vs. awareness of concept of community interventions. (e.g. , "How does it work?") represents gathering knowledge on how to use the innovation (Rogers, 1995, p. 165). (e.g., " Why does it work?") which is gathering info on the underlying function of the innovation (Rogers, 1995, p. 165). Rogers (2003) defines communication as "a process in which participants create and share information with one another in order to reach a mutual understanding" (p. 5). This process involves those who have knowledge of and experience with the innovation, those that do not, and the communication channels (interpersonal, mass media) between them. The communication channels are an important aspect in that "most individuals evaluate an innovation not on the basis of scientific research by experts but through the subjective 98 Communication behaviorinterpersonal channel (inside org) Communication behaviorinterpersonal channel (with developer) Communication behavior -mass media channel Communication behaviorscientific/local data Communication behavioropinion leader Communication barrier and facilitator -differing funding priorities Communication barrier and facilitator -communication and resource constraints evaluation of near peers who have adopted the innovation" (p. 36). Includes meetings w/boards, collaborative meetings with potential partners, other meetings that fold-in planning elements; negotiating; debating These decision-makers will often share their perceptions and initial attitude about an innovation with their peers in an attempt to reinforce their beliefs (Rogers, 1995). The amount of personal contact between the developer and potential user is important. Personal contact is the best-validated principle on knowledge transfer and organizational change (Backer, Liberman, & Kuehnel, 1986).
Individuals perceived by potential users as credible and trusted sources of information can influence the decision to adopt or reject an innovation (Barker, 2004;Dearing, 2004). For organizations and communities, opinion leaders help shape individual's opinions of and the culture around the innovation (Backer, David & Soucy, 1995). Funding priorities are established by Federal and State Departments as well as Foundations and may not align with the mission of the organization. The funding may also not budget for assistance in transferring an innovation. The communication of innovations is limited in that developers often publish new ideas, strategies or practices in scientific journals that potential users may not be aware of, may not have the time to search for, may not have access to (systemically, financially), and most likely was written with the scientific community as the audience and not in a manner that discusses the nature of implementation from the perspective of a potential user. There is often a substantial time lag between the generation of the 99 Communication barrier and facilitator -readiness for utilization Communication barrier and facilitator -differing theoretical orientations between developer and user Communication barrier and facilitator -intermediary Communication barrier and facilitator -user-oriented info Communication barrier and facilitator -participatory process Communication barrier and facilitator -incentives or rewards Persuasion -characteristics of the innovation -relative innovation and its dissemination, so much so that the innovation may be dated when the potential user becomes aware of its existence (Mills, 2002). Research suggests that once scientific validation of a program is achieved, few community practitioners utilize the information in their local community agency. Also, "the mere dissemination of articles and reports to practitioners does nothing to ensure that research-based information is actually utilized in practice settings" (Mills, 2002, p . 7). The difference in training, experiences, theoretical orientations, cultural beliefs and problem-solving approaches of researchoriented prevention scientists and practiceoriented users and/or citizen participants can serve as a barrier. Developers and potential users possess different paradigms and cultures and there is a lack of opportunities for information exchange between the two (Mills, 2002;Price, 2000). Use of outside consultants in the transfer process has been found to increase the success of adoption.
The provision of training workshops, conferences, and publications can help. It is also suggested that developers can transfer strategies to intermediary organizations as a way of reaching the community (Dearing, 2004;Mayer & Davidson II, 2000;Portnoy, Anderson, & Eriksen, 1989). May include evaluators The innovation should be described in a manner that is easily understood by potential users. This includes translating scientific information into abbreviated, understandable and accessible documents suitable for users. Potential users should be involved as early as possible in the transfer process. These can be used during the transfer process to influence the decision to adopt and with later implementation. "The degree to which an innovation is erceived as being better than the idea it  -unspecified unspecified 2003, p. 422). In this stage the organization identifies and defines a problem or need. The organization then prioritizes the problems or needs before actively searching for a solution.
Matching involves an attempt to fit the innovation with the organization's need, context and values. The organization is actively trying to determine how well the innovation would address its need while also determining how seamlessly the innovation could be implemented within the organization's structure and climate. The organization considers the feasibility of implementing the innovation as well as any potential consequences of adopting. If there is a good fit then adoption is likely to occur; a mismatch is most likely to lead to rejection. The potential user is actively seeking information about the innovation's advantages and disadvantages as they relate to the problem or need. Agreement on need -solution; problem solving Funding, sustainability Carefully thought-out plans for how the individual, organization or community will adopt the innovation in their setting, and realistic well-defined goals for the transfer process are essential to successful transfer. Identifying lead agency and other partners An optional innovation-decision is when one member of a social system makes a decision independent of other members. Collective innovation-decisions are made by consensus among members of a social system. Consensus , voting Authority innovation-decisions are when the few individuals in a system who posses the power and knowledge make the decision. Role of staff involved in making decision Decision info not fitting in other categories.