Process Evaluation in a Randomized Community Trial of a Youth-Aimed,
Substance-Use Prevention Media Campaign

Authors:Maria Leonora G. Comello, comello.1@osu.edu; Michael D. Slater, slater.59@osu.edu; Kathleen J. Kelly, kathleen.kelly@csu.edu;
Corresponding author: Maria Leonora ("Nori") Comello, comello.1@osu.edu.

(pdf version)

---

Abstract:

The article describes process evaluation in a NIDA-funded randomized community trial of a substance-use prevention media campaign called "Be Under Your Own Influence." The two-year media intervention took place in schools with some supporting community activities, and was effective in reducing uptake of marijuana and alcohol by up to 40 percent.¹ Because process evaluation and technical support are key to the effectiveness of such an intervention, the article will provide detailed description and analysis in order to guide similar interventions. We first discuss the development of the process evaluation instrument and approach, and then present campaign implementation issues that were discovered through process evaluation, and the actions taken to address these issues. Finally, we offer guidelines for other researchers who are implementing process evaluations.

Process Evaluation in a Randomized Community Trial of a Youth-Aimed,
Substance-Use Prevention Media Campaign

Process evaluation is an important task in any public health campaign. It is particularly essential during longitudinal and multi-site field studies, so that research staff can determine whether the intervention is being implemented as intended, and therefore whether implementation is consistent across sites over the course of the study.²Moreover, if the campaign is successful, process evaluation provides an account of the implementation process to guide those who may wish to replicate the campaign. Unfortunately, despite appeals from public health and communication researchers for more information on process evaluation, there is a scarcity of such reports.³

This article contributes to public health knowledge by describing the process evaluation of a large-scale randomized community trial of a substance abuse prevention campaign aimed at youth. The key intervention was a two-year, school-based media campaign called, "Be Under Your Own Influence," which was successful at reducing the uptake of marijuana and alcohol among middle-school youth.¹ This case study describes the development and administration of the process evaluation instrument, issues that were discovered through process evaluation, and actions taken as a result. The "lessons learned" section suggests ways to address the challenges of conducting field research, in terms of balancing the requirements of research and meeting the needs of target communities.

Background

The early initiation of substance use remains widespread among American adolescents. According to Monitoring the Future data (an ongoing University of Michigan study of the behaviors, attitudes, and values of American secondary school students, college students, and young adults), 22 percent of eighth graders reported having used an illicit drug, 28 percent reported having used cigarettes, and 44 percent reported having used alcohol (lifetime use figures).⁴ Given evidence that early initiation is associated with a variety of negative outcomes,^5,6 a focus on prevention among early adolescents is a key public-health goal.

However, the success of media-based campaigns in preventing substance use has been mixed. For example, the most recent evaluation of the National Youth Anti-Drug Media Campaign showed little correlation between campaign exposure and youth behavior change, and, in fact, suggested that there were negative effects on perceived norms.⁷ Other researchers have examined the effectiveness of school-based prevention curricula in conjunction with environmental-change strategies (such as media campaigns). A review of these studies found little evidence that environmental change strategies (either in conjunction with prevention curricula or alone) had an impact on behavior.⁸

In light of the mixed success of environmental change campaigns, the "Be Under Your Own Influence" campaign is noteworthy because it was more effective in reducing odds of uptake relative to a prevention curriculum in a carefully evaluated study. As reported by Slater and colleagues,¹ the study tested the impact of the campaign (in combination with a community-based media effort) on the uptake of marijuana, alcohol, and tobacco among middle-school students. Eight media treatment and eight control communities throughout the U.S. were randomly assigned to a condition. Within both media treatment and media control communities, one school received a research-based prevention curriculum and one school did not, resulting in a crossed, split-plot design. Four waves of longitudinal data were collected over two years in each school. Youth in intervention communities (N = 4,216) showed fewer users at final post-test for marijuana [odds ratio (OR) = 0.50, P = 0.019], alcohol (OR = 0.40, P = 0.009) and cigarettes (OR = 0.49, P = 0.039), using one-tailed tests. Growth trajectory results were significant for marijuana (P = 0.040), marginal for alcohol (P = 0.051) and non-significant for cigarettes (P = 0.114).

The media intervention took place over two school years in treatment communities. The tangible products of the school-based intervention included two sets of four posters (one set for each year) bearing the "Be Under Your Own Influence" theme. Figure 1 presents poster samples from the first year of the campaign. (For more information on the development of the campaign, see Kelly, Comello, and Slater.⁹) In addition, schools were given supplementary materials developed specifically for the campaign and intended for youth, such as book covers, banners, tray liners, T-shirts, water bottles, rulers, lanyards, and stickers. We provided these materials to participating schools at the beginning of each school year, with the expectation that school staff would display the posters in highly visible locations throughout the school and distribute the promotional items on appropriate occasions throughout the year.

Schools served as the main distribution channel for these materials because of their potential to reach target audience members (youth) where they spend most of their time - at school. Maximizing this distribution channel was central to implementation, given the importance of exposure to the success of health campaigns.¹⁰Research staff members, however, were not on site and therefore had to depend on school contacts to distribute materials appropriately. Process evaluation provided a way to track distribution of products within schools and, perhaps of equal importance, to build-in opportunities to dialogue with school contacts and reinforce their partnership with the research staff. In most cases, contacts were the safe and drug-free program coordinators at the schools, and were involved in both school and community affairs.

Although schools served as the main distribution point for most of the tangible campaign materials, we also provided certain community members with community mobilization training, in recognition that community efforts could reinforce the school-based intervention.¹ We scheduled a full-day community workshop at the beginning of the first school year to introduce the project to key school and community personnel. Community personnel included those who were involved in prevention and/or youth activities, such as youth group organization leaders, law enforcement officials, and health coalition members. Using principles from the Community Readiness Model,¹¹,¹² the workshop was intended to help participants assess their community's readiness to take action on the issue of youth substance abuse prevention. Whenever possible, the workshop was held in conjunction with local prevention coalition meetings, and participants discussed prevention strategies that would best use their resources. There were no additional required workshops throughout the project, although the research staff was available to provide technical support and materials when requested.

To support community efforts, each community contact received a "media kit," which contained press releases about the campaign and general articles about youth substance abuse prevention that could be used in local print media. Furthermore, contacts were given radio and television public service announcements that could be used in local broadcast media. Again, because research staff members were not on site to ensure distribution, we relied on process evaluation and partnerships with field contacts to monitor how these materials were being used.

A key concern in process evaluation is specifying complete and acceptable delivery of an intervention.² Given the field environment, we expected to see variation in how schools and communities chose to implement materials; therefore, we did not expect rigid adherence to a set of instructions. It would have been artificial and counterproductive, for example, to restrict the use of promotional items at a school's event simply to ensure consistency with other schools that may not have such an event. Treatment communities were provided, however, with in-person training (in conjunction with the community workshop) and a handbook, which specified guidelines for implementation, expected uses of materials, and other ideas to support prevention in the school and the community.

Development and Administration of the Process Evaluation Instrument

Consistent with the recommendations of Saunders and colleagues,² process-evaluation questions were organized by intervention component and followed basic principles when relevant: intervention fidelity (extent to which the intervention was implemented as planned), dose (extent to which components were delivered, received, and enjoyed), and context (organizational and other field issues that affected implementation). Our approach was to first present contacts with questions about community efforts, followed by questions about media efforts, and ending with questions assessing the overall progress of the project (see Table 1). It should be noted that the instrument did not include process evaluation items for the curriculum component of the intervention. This was because the curriculum instructors (teachers from the school who had received training in curriculum delivery) met regularly with the instructor-trainer to assess progress and discuss any issues that arose in the classroom. Therefore, we did not believe it was necessary to further monitor curriculum activities.

Community-efforts component. The first section of the instrument asked questions about the impact of the community workshop, local efforts for substance abuse prevention, and overall community conditions. The primary goal of this section was to track the impact of the workshop and other prevention-related community efforts, in order to provide context for the overall intervention.

To develop questions for this section, research staff members drew on the Community Readiness Model, which states that a community's level of readiness to address an issue is determined by a number of dimensions, such as the presence of local efforts and resources, as well as the overall salience of the issue in the community.^11,12 Questions were designed to gather information about these factors. For example, contacts were asked whether there had been recent events in the community that may have placed a spotlight on substance-use issues. We also wanted to know whether local community efforts were gaining momentum. Questions that assessed this aspect included whether there was an active local coalition, how often it convened, and what prevention strategies the coalition was pursuing. Table 1 provides additional examples of questions in this section.

It is important to note that the Community Readiness Model includes a formal procedure for the assessment of community readiness, which consists of in-depth interviews with several community members, with responses transcribed and coded by independent raters. These assessments were performed in each community at baseline and at the end of two years, with results reported elsewhere.¹³ Due to the effort involved in conducting full assessments, they were not conducted every six months as a part of the regular process evaluation. However, the community-oriented questions in the process evaluation instrument provided a quick "snapshot" of community conditions.

Media component. In the next section of the instrument, contacts were asked about their utilization of the media campaign and supporting materials. Considering the process-evaluation principles mentioned above, message fidelity does not seem to be as important an issue in a media campaign (as it might be in a teacher-delivered curriculum, for instance), because the basic message and visuals of posters cannot be changed. More relevant process-evaluation considerations, then, are dose and context. To assess the delivered dose, the contacts were given a list of the items we had given them at the beginning of the school year (posters, all promotional items, media kit, and radio/television public service announcements), and were asked to indicate where, when, and with what audience they had used each item. To assess dose enjoyment and satisfaction, the contacts were asked which items were the most and least liked by the target audience. To assess context, contacts were asked whether any of the press releases in the media kit had been used or adapted by local print media. They were also asked whether the local media had covered the project or other prevention activities.

Wrap-up component. The instrument concluded with wrap-up questions, which were designed to elicit information about anything that might have made implementation harder or easier. As such, the questions provided more information about context and overall satisfaction with the campaign. Table 1 provides examples of questions included in this section.

Table 1. Examples of process evaluation questions by intervention component.

Question Examples	Process-evaluation issue addressed
Community efforts Is there a coregroup of people in your community who are working together on youthdrug-prevention issues? In the last sixmonths, have any critical events occurred that were directly related to youthsubstance-abuse (e.g., arrests, injuries, deaths, etc.)?  If so, what was your community's response? In the last sixmonths, have any pressing events occurred that diverted your community'sattention from youth substance-abuse? (Examples of such events would include natural disasters, urgent localissues, etc.)  If so, what was yourcommunity's response?	Context
Media Please tell us aboutany special events, activities, or projects in the last six months that arerelated to youth drug-prevention. Please tell us how you've used the following promotional items in the last six months.  (List provided, with spaces to indicate where,when, and how much used). Which items did thestudents like best or were most popular? Which items were liked least? Please tell us about any media coverage(newspaper articles, columns, photos, etc.) of youth drug-preventionactivities/issues in the last six months. Of all the communication "tools" we've sentyou (media manual, promotional items, public service announcements, presskits with articles/columns, etc.), which do you think has been mostuseful? Leastuseful? Why?	Dose delivered,satisfaction, context
Wrap-Up How have you integrated this project withother prevention activities and projects that are already going on in yourschool and community? In your view, what has been the mostchallenging aspect of working on this project? How will your participation in this project affect future prevention efforts in the school/community? Do you have any other comments or suggestions?	Context,satisfaction

The process evaluation was conducted every six months over the two-year implementation period. When a community was due for process evaluation, we mailed the contact a survey, accompanied by a cover letter that gave a brief explanation of the purpose of the survey, the dates of their previous and next surveys (if applicable), instructions for returning, as well as a copy of their last completed process evaluation instrument (if available). We provided each contact with options for responding to the survey: by mail, phone, or e-mail. When phone conversations took place, the staff member conducting the evaluation documented the responses to the survey. Project investigators and other key staff members reviewed and discussed completed surveys to coordinate any further action, if needed.

Findings and Resulting Actions

Evaluation completion. Process evaluations were conducted with contacts from all eight of the treatment communities throughout the duration of the project. The evaluations occurred no later than a few weeks past the target dates, although in most cases, reminders had to be sent or the evaluations had to be completed by phone. Overall, the phone conversations yielded the most detailed information and served as the best opportunity to build relationships with contacts. Even when the evaluation was completed by phone, the evaluation instrument was used to provide structure to the conversations, and (when reviewed by participants in advance) the instrument encouraged participants to provide more thoughtful responses.

Although the process evaluation proceeded smoothly in most of the communities, there was one community contact that did not respond after several reminders. This prompted staff members to make a site visit to investigate further. After meeting with the contact, it was discovered that school administrators had forced the contact to assume responsibility for the project, and the contact felt some resentment towards the extra duties. To ease the workload, it was decided that the contact would split some duties with another staff member. This change enabled staff members to work with the community contact to implement the intervention.

Community efforts component. Overall, the section on community efforts was helpful in assuring research staff that extraordinary events did not occur in a specific community that would have negatively impacted that particular community's readiness to address youth substance use. In response to the question about local events that may have impacted substance use, the most commonly reported events were arrests for drug possession. Responses were similar across communities. It was the case, though, that some communities seemed more active than others in addressing youth substance-abuse prevention, as indicated by responses to questions about local coalitions or groups.

Media component. In the media component evaluation, all of the contacts reported that they had used the posters in schools, and in some cases throughout the community, which was reassuring and consistent with project expectations. Schools varied in their usage of promotional items. A few items (such as tray liners and book covers) were not used in some locations, because schools did not need these items. This prompted research staff members to more carefully assess school needs and to develop a "menu" of items from which contacts could choose the items they were most likely to use.

The process evaluation did not provide much information regarding the schools' utilization of the radio and television public service announcements (PSAs). Specifically, the evaluation instrument asked whether the PSAs had been used, and if so, how frequently they had been broadcast, and whether proformas were available. Most respondents left this area blank, and follow-up interviews with some contacts revealed that there was a lack of local broadcast outlets in their communities. This prompted staff members to reevaluate the usability of these supplementary items for future projects.

Wrap-up component. Overwhelmingly, contacts reported that the major challenge to implementing the intervention was a lack of time. Contacts also reported that getting the word out and maintaining a committed group of people to work on the project were challenges to implementation. Furthermore, one community contact reported that the school had not achieved a good score on an academic standardized assessment, and that, as a result, extracurricular activities (including school-based prevention efforts) were in danger of being cut to allow students and staff to focus on academic priorities. This finding prompted staff members to talk with the contact and gather materials that the contact could present to school administrators, if necessary, which emphasized the association between academic achievement and non-use of substances. In so doing, we sought to demonstrate the role of this prevention intervention in supporting overall educational goals. The project was not cut.

Another key issue that arose for community contacts was integrating the project into existing prevention efforts. In one community, this was difficult because the community had a pre-existing, strongly-branded youth-aimed health campaign, which was run by a local unit of a national organization. Research staff did their best to assure the contact in this community that the campaigns could co-exist, and that they wholeheartedly supported local efforts, given the potential for mutual reinforcement. The school did use some of the project's materials, but overall the materials were probably underutilized in the school and community because of perceived competition. This situation prompted staff members to be more aware that this situation may arise in other communities, and to emphasize to contacts that the goal was not to displace, but to complement, existing activities.

This section also asked about new ways to use media/promotional items and requests for assistance. One community contact shared with research staff the idea of distributing wallet cards to 8th graders, bearing the "Be Under Your Own Influence" logo, as part of the 8th grade graduation ceremony. Other community contacts requested information aimed at parents, which explained the signs of substance use. This information could be distributed through local media, parent newsletters, or other channels in order to facilitate parent-child discussions. This prompted staff members to gather research-based materials, and to provide them to all of the treatment communities, so that they could be used to supplement existing materials.

Strengths, Limitations, and Lessons Learned

When analyzing process evaluation approaches, it is useful to examine the extent to which the evaluation assesses program fidelity, dosage, and contextual issues.2 One strength of this intervention (in terms of fidelity) is that intervention communities were provided with a day of in-person training, as well as a comprehensive handbook on the expected implementation of the campaign. By tracking how materials were used, the process evaluation confirmed that materials were being used appropriately.

In terms of dosage, the process evaluation assured research staff that media exposure occurred at a basic level. It is important to note, though, that distribution numbers for posters and other materials were based on estimates, rather than verifiable numbers. This made it difficult to calculate exposures for various distribution channels, as might be done in other process evaluation approaches. Consequently, we were not able to use exposure information from the process evaluation to cross check with self-reported exposure measures on youth surveys. Also, as previously noted, an inherent challenge of school- and community-based media campaigns is that it is difficult to spell out "complete and acceptable delivery" of the intervention, due to the variability of field environments. For future projects, it would be better to refine the conceptualization of successful implementation and develop ways to communicate this to contacts so that they can more easily envision what is expected of them. Providing contacts with a short narrative about a similar school that successfully implemented the campaign might achieve this goal.

Another aspect of dosage assessed in process evaluation is the extent to which the target audience enjoys and feels satisfied with the intervention.² Enjoyment and satisfaction were assessed in the field by asking contacts which materials were most and least liked. Although feedback from youth was solicited during pretesting of the materials, youth in the treatment communities were not directly asked about their reactions to the materials while the study was underway. The process evaluation would have been stronger if research staff had sought feedback from youth and other stakeholder groups (e.g., teachers, parents, local media representatives), who were exposed to the campaign in the field. Future projects will seek to include greater diversity in respondents.

In terms of contextual factors, another strength of this process evaluation was that it sought information about community-level activities, including media coverage of the intervention, as well as about the integration of the intervention into the overall school and community environment. The evaluation would have been stronger if it had sought feedback from representatives of other stakeholder groups. In some cases, contacts were not able to provide substantive responses on the community-focused section of the survey if they were not actively involved in community activities themselves.

Furthermore, the process evaluation did not ask for detailed information about the implementation of the in-school prevention curriculum. Findings from the overall study showed that the media intervention, combined with community efforts, was more effective than the curriculum.¹ However, the process evaluation could not help to explain this finding, as it did not monitor the curriculum implementation. A better integration of process evaluation data for all components would have enhanced the evaluation's ability to interpret results.

An additional dimension to consider when analyzing strengths and weaknesses of process evaluation is the extent to which the process evaluation itself was successfully implemented. This process evaluation was successfully implemented, and research staff members were able to conduct evaluations every six months with all target communities, as intended. Responses gathered through phone interviews provided the richest information, although the knowledge gained from these interviews might have been even richer if they had been conducted in person. Although conducting regular on-site visits was not feasible given the resource constraints of this project, in future efforts, it is important to structure project activities so that site visits include process evaluation whenever possible.

It seems that many of the challenges faced during the process evaluation arose due to geographical distance, the inability to make regular site visits, and consequently, a heavy reliance on contacts within each community. Given this constraint, the regular communication required by process evaluation helped to strengthen relationships with community contacts. The relationship-building aspect of process evaluation helped strengthen the partnerships between research staff and community contacts, and helped both groups orient themselves toward shared goals. Periodic evaluation also allowed us to provide additional technical assistance when appropriate, and to share resources developed for one community with all other treatment communities.

The main lesson learned through this process evaluation was that it is a struggle to conduct any research under real-world conditions, but it is this struggle that makes process evaluation a necessary task. Based on our experiences, we offer the following guidelines to others who must conduct process evaluation in field environments:

• Introduce the concept of process evaluation early. Some field contacts may not be familiar with the term "process evaluation," and may view it as an unnecessary burden. Therefore, it may help to provide potential participants with the rationale for conducting the process evaluation, the expected time commitment, and a tentative schedule, at the beginning of the project (or even during the recruitment stage). Research staff may choose to call the process evaluation "progress checks" (or something similar), which are conducted periodically to ensure that each site is receiving the support it needs.
• If feasible, designate one person who is familiar with project components and who ideally has had some personal contact with communities to conduct the process evaluation. Having one person may help ensure consistency within each community and across treatment communities. If this is not possible, all staff members involved in process evaluation should follow similar procedures and should communicate activities with each other in order to ensure consistency.
• Become familiar with the calendar at each specific field site to guide the selection of dates for sending evaluations and conducting follow-ups, as well as to see if there are upcoming events that can dovetail with program goals. Calendars and list of upcoming events are often posted on organizations' websites.
• To encourage prompt response, ensure that process evaluation "packages" are easy to access and process, and consider sending reminders if necessary. As suggested by Dillman,^14,15 important components of the package include: a personalized cover letter, a real signature on the cover letter, information on multiple ways of responding, survey questions that are easy to interpret, and delivery via first-class mail. If contacts do not return their surveys within a reasonable amount of time, a reminder can be sent via e-mail (with the survey attached), or by phone.
• Even if contacts return surveys on time, plan on discussing the survey with them. Conversations can serve multiple goals, chief among them gathering information and building relationships. Conversations can even be conducted with those contacts who return the instrument by the target date. Conversations can also be initiated to follow-up on responses that require clarification. In this manner, the instrument serves as a springboard for further discussion and leads to a better understanding of field conditions.
• Ask contacts to prepare for evaluations by keeping an ongoing record of intervention activities. It is a complex task to track the use of a multitude of materials (such as those offered to communities in this study). Therefore, research staff in similar projects might want to ask participants to keep an ongoing log of material usage. In this study, participants were asked to keep a file with notes on media material and promotional item usage, as well as any newspaper clippings of media coverage. We also suggest giving contacts the process evaluation form in advance so that they may record activities directly on the form.
• Share ideas gleaned from the process evaluation with key staff, as well as other treatment communities, as appropriate. As discussed in the previous section, some communities developed ideas that enhanced project implementation and the utilization of materials, and we chose to share these ideas with other communities. Although some might argue that sharing ideas confounds treatment effects, we argue that it helped equalize conditions among communities, thus allowing for better evaluation of intervention effects.

In relation to confounding effects, other questions can be posed: does the process evaluation itself confound treatment effects? Are field contacts more motivated to implement the intervention if they know their efforts will be evaluated periodically? Does field implementation vary depending on contacts' relationships with process evaluation research staff? We would argue that these issues reflect the realities of working in partnership with schools and communities on research projects. In most field research, there is a trade off between the control that researchers have over experimental conditions and the ecological validity conferred by real-world conditions. Precisely because field conditions vary, process evaluations are important to ensure that interventions are implemented at a basic level, and that there is a means of tracking substantive differences in community conditions. Therefore, it is our view that the benefits of process evaluations far outweigh any potential confounding effects they may have on interventions.

In summary, process evaluation enhanced the implementation of the "Be Under Your Own Influence" campaign, which was later demonstrated to be successful in reducing uptake of substances among middle-school youth. The findings reported here point to challenges that may arise in school and community environments, as well as opportunities for resource sharing and relationship building. Moreover, the guidelines provide researchers with suggestions for facilitating process evaluation in field research. Finally, as resource sharing was important to the quality and consistency of implementation among treatment communities, we suggest that information sharing among field researchers is also essential. Thus, our final suggestion is that others involved in community- and school-based health research also report process evaluation results so that practices can be shared with all.

Acknowledgments

This work was conducted as part of a grant from the National Institute on Drug Abuse (DA12360) to the second author. We wish to thank our field contacts for their support of the process evaluation and of the study as a whole.

---

References

1. Slater MD, Kelly KJ, Edwards RW, et al. Combining in-school and community-based media efforts: Reducing marijuana and alcohol uptake among younger adolescents. Health Educ Res. 2006;21(1):157-167.
2. Saunders RP, Evans MH, Joshi P. Developing a process-evaluation plan for assessing health promotion program implementation: A how-to guide. Health Promot Pract. 2005;6:134-147.
3. Noar, SM. A 10-year retrospective of research in health mass media campaigns: Where do we go from here? J Health Commun. 2006;11:21-42.
4. Johnston LD, O'Malley PM, Bachman JG, Schulenberg JE. Monitoring the Future national results on adolescent drug use, 2004, Table 1: Trends in lifetime prevalence of use of various drugs for eighth, tenth, and twelfth graders. Available at: http://monitoringthefuture.org/data/04data/ pr04t1.pdf. Accessed June 1, 2005.
5. Anthony JC, Petronis KR. Early-onset drug use and risk of later drug problems. Drug Alcohol Depend. 1995 Nov;40(1):9-15.
6. Lynskey MT, Heath AC, Bucholz KK, et al. Escalation of drug use in early-onset cannabis users vs co-twin controls. JAMA. 2003;289(4):427-33.
7. Orwin R, Cadell D, Chu A, et al. Evaluation of the National Youth Anti-Drug Media Campaign: 2004 report of findings: Executive summary. August 2006. Available at: http://www.nida.nih.gov/DESPR/Westat/NSPY2004Report/ExecSumVolume.pdf. Accessed January 1, 2007.
8. Flay B. Approaches to substance use prevention utilizing school curriculum plus social environment change. Addict Behav. 2000;25(6):861-85.
9. Kelly KJ, Comello MLG, Slater MD. Development of an aspirational campaign to prevent youth substance use: "Be Under Your Own Influence". Soc Mar Q. 2006;12(2):14-27.
10. Hornik R. Exposure theory and evidence about all the ways it matters. Soc Mar Q. 2002;8(3):30-7.
11. Oetting E, Donnermeyer J, Plested B, Edwards R, Kelly K, Beauvais F. Assessing community readiness for prevention. Int J Addict. 1995;30(6):659-83.
12. Plested BA, Edwards RW, Jumper-Thurman P. The Community Readiness Handbook. Fort Collins, CO: Tri-Ethnic Center for Prevention Research. 2003.
13. Slater MD, Edwards RW, Plested BA, et al. Using community readiness key informant assessments in a randomized group prevention trial: Impact of a participatory community-media intervention. J Community Health. 2005;30(1):39-53.
14. Dillman DA. Mail and telephone survey: The total design method. New York: Wiley. 1978.
15. Dillman DA. Mail and internet surveys: The tailored design method. New York: Wiley. 2000.