People
The first P, people, had been clearly determined prior to the start of the study—parents who were the primary household food gatekeepers, were 20–45 years old, lived in the catchment area (Arizona or New Jersey), and had at least one child aged 2 to < 6 years, at least basic English or Spanish reading skills, and regular Internet access. Parents were the primary target, but their preschool children and family unit also were considered “customers.” Thus, it was critical to be certain all of them gained satisfaction from the “product” (a healthier, happier family delivered via the HomeStyles program).
Project staff also were part of the people mix because they were responsible for establishing and maintaining relationships with parents, either remotely through the program website and printed materials or in actuality via phone, SMS (texting), and email. To support recruitment and retention [
41‐
44], bilingual, culturally sensitive project staff (HomeStyles specialists) were trained in customer service. The specialists were trained to rapidly respond to participant queries coming in by email or the dedicated toll-free line and use a positive, nonjudgmental, courteous, “can do” tone. Scripted responses were created to ensure accuracy and consistency of information proffered and equal handling across treatment groups.
Product
Product, the second P, was a major component of this overall project, taking over 2 years to create the core products (i.e., instructional materials). The development of HomeStyles is described in detail elsewhere [
2‐
6]; a brief summary follows to provide context for this paper. HomeStyles is an online program designed to enable and motivate parents of preschool children to make quick, easy, no-cost childhood obesity-preventive changes to their home environments and family lifestyles. This program has a social ecological framework [
45], is grounded in Social Cognitive Theory [
46,
47] and Adult Learning Theory [
48‐
53], and uses motivational interviewing principles [
54‐
57]. Content was developed to be responsive to the latest research findings and childhood obesity prevention recommendations [
1,
58‐
61]. All instructional and data collection materials were thoroughly pretested with members of the target audience to ensure a high degree of comprehension, relevance, usability, interest, preference, attractiveness, and satisfaction [
2‐
5,
62‐
64].
HomeStyles materials and study procedures were methodically and carefully developed with input and buy-in from parents at every step because considerable evidence suggests participant perceptions of program characteristics (e.g., content, packaging, promotion, time commitments) affect whether individuals choose to enroll and continue in a study [
65]. A common reason families do not participate in studies is that they dislike the idea of being the subject of research [
65]. Accordingly, care was taken to avoid using terms such as “subject” (“parents,” “families,” or “participants” were substituted) and “research” (“study,” “program,” or “project” were used instead) as well as terms implying “schoolwork” (e.g., instead of “work on a lesson,” participants “get to review a guide,” or instead of “take the posttest,” parents “get to go to the survey café”). Treatment groups were not referred to as “control” or “experimental” (rather, “safe HomeStyles” and “healthy HomeStyles”), because it was important to keep participants blind to their treatment group, as well as to avoid research connotations, disappointment that may occur with assignment to the control group, efforts by the control group to seek alternative access to experimental group materials, and/or differential dropout rates [
66‐
79]. This method of referring to treatment groups also served as a constant reminder to staff that this RCT had two analogous treatments differing only in content.
Parents who completed an online screener and met eligibility criteria were immediately invited to participate and complete the informed consent. To facilitate recruitment, the informed consent was visually appealing, clearly written using common language, and easy to complete and submit [
23,
80‐
82]. Participants who consented and finished the baseline survey were systematically assigned by computer to a treatment group. Participants could immediately access instructional materials after group assignment with the aim of limiting attrition [
83]. Experimental group parents received web-delivered instructional materials focused on weight management-related topics (i.e., diet, physical activity, sleep) and the attention control group received a bona fide intervention with materials covering home safety topics. Both treatment groups had access to a series of 12 instructional guides in a 4-page minimagazine format. All parents received the same first guide (specific to their treatment group), which provided an overview of the intervention and tips for deciding which subsequent guides would be the best match for their families’ goals. With the exception of the first guide, parents could select any guide in any sequence. Progress through the RCT was as follows:
-
Level 1: Baseline survey
-
Level 2: Receive overview guide and then every ~ 16–30 days, choose a different guide for a total of four guides, then complete the midpoint survey
-
Level 3: Every ~ 16–30 days, choose a different guide for a total of four guides, then complete the post survey
-
Level 4: Choose a different guide or revisit a previously chosen guide, and ~ 30–60 days later, complete the follow-up survey
-
Level 5: Choose a different guide or revisit a previously chosen guide, and ~ 30–60 days later, complete the long-term follow-up survey
Completion of the entire RCT was estimated to take 12–18 months [
6].
Considerable effort also was spent on creating study surveys that were psychometrically sound as well as accurately understood, enjoyable, and easy to complete by the target population [
6,
62‐
64,
84‐
86]. Research shows that participant enjoyment of data collection procedures, especially the first round of data collection, contributes favorably to participating in future data collection [
41,
85,
87]. Thus, all HomeStyles RCT data collection was conducted online at a time of the participant’s choosing. Survey design strategies to help reduce participant burden included arranging items with the same answer choices (e.g., strongly agree to strongly disagree) in a matrix with alternating shading to increase reading ease as well as speed and accuracy of completion; using font treatments (e.g., bold, color, italics) for emphasis to promote rapid comprehension; including brief, clear instructions for completion; varying clip art page headers and question types to promote interest; limiting the number of items appearing at one time to minimize the need to scroll the page; using primarily radio buttons and drop-down boxes with few items that required free text answers; using adequate white space to decrease visual burden and give a feeling of rapid progress; and giving mental breaks to prevent fatigue (e.g., periodic placement of a photo evoking relaxation and calmness, such as a beach, with a caption encouraging the participant to take a deep breath and think about the fresh scent of the ocean). In addition, the online surveys were thoroughly pretested and refined, then pilot-tested and refined to ensure there were no technical issues that would frustrate participants. Pilot-test participants (
n = 550) rated overall satisfaction with the survey as 4.5 out of 5 points.
Supplementary products were developed to enhance the value of the core products. These included tracking forms coordinated with instructional guides that promoted goal setting and monitoring of progress toward goals, mailed enhancements that facilitated application of the guide content (e.g., measuring cups to support portion control, cutting board to encourage fruit/vegetable intake), and extra guide-specific information and resources on the website (e.g., tips, goal ideas, links to helpful external videos and websites).
Supporting services offered were the friendly HomeStyles specialists who were readily available by toll-free phone or email. The services offered included technical problem solving (e.g., website login), providing additional facilitative information, and encouragement.
Price
Price, the fifth P, centers on “costs” to the participants. Although there was no dollar cost for parents associated with participation, there was a time cost as well as the potential cost of stress on family relationships associated with participation (e.g., stress associated with devoting time to the intervention that would otherwise be used differently, setting and working toward goals, and coping with possible family resistance to change). Time commitments are an important determinant of participation [
65,
95]. To address time costs, recruitment materials stated there were 12 guides that each took about 15 minutes to review, that families would spend a few minutes daily making simple changes to help their families, and that total length of time for the intervention was about 12 months. Materials also indicated the changes were easy to make and could fit into busy schedules and tight budgets. Additionally, recruitment materials and reminders to participants stated that parents would receive stipends for completion of study surveys as well as gifts to help them implement simple changes.
Stipends were offered for two reasons: to compensate participants for the time spent completing surveys (i.e., working for the researchers) and because stipends are associated with improved recruitment and retention success [
36,
87,
92,
96‐
100]. Cash stipends were paid after completion of each survey, and as commonly done in other studies, amounts increased modestly with each subsequent survey [
41]. “Gifts” took the form of guide enhancements (e.g., measuring cups) that not only served to facilitate application of guide concepts but also helped to forge and renew relationships between participants and researchers, gently remind parents about participation in HomeStyles, and build goodwill. Other tokens of appreciation and reminders to participate included holiday cards, refrigerator magnets, and key chains, all displaying the HomeStyles logo. Anecdotal evidence of participant appreciation of stipends and gifts included unsolicited emails from parents that told project staff how they planned to use stipends to benefit their families and about the excitement of their children when packages with the project logo arrived in the mail.
To promote frequent visits to the website and introduce an element of fun and, thus, promote retention [
41], parents had the opportunity approximately every 10 days to earn a “bonus buck” ($1 US) that would be added to their next stipend. The bonus bucks asked parents to answer an interesting or fun question (“If you wrote a song about HomeStyles, what would you call it? Who would you get to sing it?”; “CNN is on the phone—what would you tell them about how HomeStyles helped your kids?”)
To lower stress costs, recruitment announcements specified that a friendly HomeStyles specialist was available by phone or email. There was a “stress busters” and “confidence builders” section placed outside the secure login area of the project website so that potential participants could review it while deciding whether to participate. Similarly, to minimize possible relationship costs parents might encounter from potentially resistant family members, a “get more” section of the project website outside the login area encouraged parents to make this a family project and get everyone involved.
To promote retention [
42,
92,
101], throughout the intervention parents received periodic reminders about the benefits and return on their time investment (e.g., happier, healthier families); stress reduction suggestions (e.g., “make it a family agreement, not a family argument,” “if you get off track with your goals, just start over,” “choose another goal that may be easier”); encouragement to persevere (“keep yourself moving by thinking about how much you love your family and the steps you can take to keep them healthy”); and friendly, encouraging nudges delivered by email, voice mail, and/or text (per parent preference). When parents were eligible to complete a survey, friendly announcements reminded them of the opportunity to earn the associated cash stipend.
Another aspect of price for RCTs are expenses associated with recruitment and retention of participants. These included monetary costs (e.g., printing and distribution of recruitment and other study materials), staff costs (e.g., energy and time allocated to training and recruitment), and relationship costs (e.g., effort to maintain relationships with existing community partners and attract new ones). An additional cost was the stress costs associated with the intensity of recruitment and retention activities, slow accrual of enrolled participants, and loss to follow-up despite intensive staff efforts. To keep these costs in check, the research team carefully planned recruitment and retention activities and identified an array of methods to keep monetary costs under control. For example, professionally designed recruitment materials were distributed in print form only in high traffic areas likely to reach the target audience (e.g., pediatrician’s offices; child care centers; Women, Infants, and Children program [WIC] offices), paid advertising was kept to a minimum, in-kind marketing opportunities were sought (e.g., radio interviews, links from other websites), and the bulk of recruitment occurred electronically.
Project staff stress was managed by actively involving staff in recruitment and retention decisions; keeping communication open and positive; clearly communicating procedures and adjusting them as needed to be responsive to pertinent events, observations, and opportunities; rotating duties; keeping everyone up-to-date on progress and complimenting efforts; holding refresher trainings; giving staff feedback and recognition; holding occasional staff appreciation events; and maintaining a high level of staff enthusiasm for the project. These efforts fostered cohesion and a strong team culture, which paid off, as evidenced by a continual positive workplace atmosphere and low staff turnover (primarily students who graduated).
The sixth P, promotion, addressed all forms of marketing HomeStyles to prospective participants. The content of the recruitment marketing materials was informed by qualitative data collection activities (i.e., focus groups [
n = 139] and cognitive testing of intervention materials [
n = 512]) conducted as part of the overall project formative research and supplemented with quantitative preference surveys with English- and Spanish-speaking parents of preschool children residing in New Jersey and Arizona [
2‐
5]. Salient findings from the formative research relevant to recruitment materials included parents’ strong dislike of the terms “obesity” or “overweight” and messaging that implied a need to organize or “get things under control” [
3,
4]. Similarly, others have reported that parents tend to have little concern or interest related to obesity [
23,
33].
Parents preferred messages that projected happiness, fun, and quick and unique solutions to everyday challenges. They liked attention-catching colors, appreciated photos of families of varied races/ethnicities, were curious about other families’ behaviors and opinions for self-comparison purposes, valued other parents’ endorsement of HomeStyles, and had a robust desire to build stronger bonds with their children. In fact, their desire for happier, closer family relationships ranked higher in importance than improving the health of their children [
5]. Thus, the content and design of all HomeStyles materials, including recruitment materials, aimed to appeal to parent preferences. Recruitment materials were cognitively tested with both English- and Spanish-speaking parents and parent educators to confirm the clarity of the content, the appeal of the content and design, and the likelihood that parents would respond by visiting the website to sign up for or learn more about the study [
2].
Recruitment materials took a wide array of forms and were simultaneously disseminated in a variety of ways to extend the reach and speed the accrual of participants [
22,
36,
83,
102]. Nearly all recruitment was done in a passive form (e.g., posters, flyers) to contain costs and because this method was well suited to an online study. Additionally, the limited research comparing active with passive recruitment indicates that either form can work, and participants recruited by these different methods usually do not differ by baseline demographics, psychosocial variables, or attrition [
103].
Passive marketing efforts included distribution or display of printed materials (e.g., flyers, bookmarks, posters, magnets, key chains) by community partners (e.g., pediatricians’ offices, fitness centers, schools, preschools/day care centers, workplaces, community centers, health fairs, and farmers markets). To help existing community partners learn more about HomeStyles and to build new partnerships, project staff directly marketed HomeStyles to them via webinars held at convenient times, in-person visits to partners’ offices, brief talks at professional meetings, and short informational YouTube videos specifically tailored to key recruitment partners (i.e., registered dietitians, pediatricians, and early childhood educators). To ensure that community partners remembered HomeStyles, researchers made at least two personal follow-up calls or contacts to encourage them to promote HomeStyles to their clients, confirm receipt of the recruitment materials sent, and answer questions.
For electronic recruitment announcement distribution, an inventory was created and regularly updated that included addresses of listservs that could reach targeted participants directly or through trusted sources such as workplaces, religious groups, philanthropic or community organizations, preschools/day care centers, professional associations, and extracurricular activity groups. Listserv administrators were contacted by email and/or phone to encourage them to forward the electronic recruitment announcement to listserv recipients. Administrators had the opportunity to contact researchers via email or phone to learn more about the project. Whenever possible, supervisors of the listserv administrators were contacted to gain their endorsement for forwarding the emails. Listserv administrators were contacted three or more times over the recruitment period. Appeals to listserv administrators were sent from an official university email address and included university logos to establish credibility. A professional study recruitment agency also was employed to distribute electronic announcements about HomeStyles to the members of their research panel.
Other electronic recruitment venues included notices posted to websites that target parents, online local newspapers, local businesses, and parenting-related blogs. Social media also was used for recruitment, as recommended by others [
104], although it was not possible to create online communities for parents to interact with each other, owing to the need to maintain participant blinding to treatment group, prevent contamination (i.e., sharing of knowledge with parents in the other treatment group), and protect participant privacy. By adjusting security and privacy settings, it was possible to create Facebook® (Menlo Park, CA, USA) and Pinterest® (Cold Brew Labs, San Francisco, CA, USA) pages for parents to learn about the study and link these pages to the study website to facilitate enrollment.
Print media provided gratis was used as a recruitment tool, though in a limited way. A university-based collaboration with a children’s cooking magazine afforded an opportunity to include an announcement about HomeStyles in copies distributed in New Jersey. There was some success with product placements in printed newspapers, which involved mostly interviews that led to brief articles about the project with information on how to sign up. Owing to lack of access and budgetary limitations, recruitment efforts using radio and television were limited to brief interviews.
Paid advertising was purchased through Facebook® because of its widespread use and potential to reach parents. A series of 64 ads were created, all of which included a photo of a preschool child and text promoting HomeStyles. The ads systematically varied the photo (four races/ethnicities, two sexes), salutation (Hey Dads!, Hey Moms!, Hey Moms & Dads!, and Hey Parents!), and closing statement (Click here! and Find out how!). Four evaluators unanimously agreed that all eight photos had similar lighting and clarity and depicted children of a specific race/ethnicity, sex, and age who directly faced the camera and had a “happy” expression. Facebook® criteria were set to target ads to parents aged 20–45 years with children aged 0–12 years who were interested in happy kids or fitness and wellness. These ads ran for 250 days [
105].
Some direct, active recruitment marketing was conducted at community events, parent resource centers, WIC offices, and farmers markets. Internal marketing efforts overlapped with project staff stress management procedures previously described. In addition, another important component of the recruitment campaign was to exhaustively rally project staff’s families, colleagues, friends, and neighbors to distribute recruitment materials and share the word about HomeStyles. In addition, these individuals contributed ideas for new recruitment methods and opportunities.
To keep participants blind to their treatment group assignment, it was critical for all recruitment materials to be applicable to both groups. To achieve this, the study expectations for the experimental and action control group were held constant. In addition, the program was described as one that would help kids be happier, healthier, and safer—terms that were applicable and true for both treatment groups.
Process
The final marketing P, process, is focused on ensuring standardized, customized, and efficient service delivery. The website helped to ensure that service delivery would be standard across participants. Because not all transactions were web-based, standard operating procedures (SOP) and manuals were created and implemented to promote quality control and ensure all staff performed them uniformly. These SOPs included templates (e.g., scripts, email text) for responding to commonly asked participant questions, procedures for handling technical problems, and complete and illustrated instructions for preparing all participant mailings (e.g., enhancements, holiday cards). In addition, staff had a clear chain of command for guidance for handling problems that did not have an SOP or when they felt the typical SOP needed to be customized to better meet a participant’s needs.
Timely product delivery and responses to parent queries were top priorities. To address this and improve retention [
38], the project had a dedicated email account for participants and a toll-free phone line that was staffed consistently and from which calls were responded to quickly. Given that providing reinforcement soon after completing RCT activities helps promote retention [
41], parents were sent their stipends electronically in the form of a gift card for a store of their choosing within 2–3 days of completing a survey. Likewise, soon after selecting a new guide, staff mailed a printed copy of the guide and, for at least every other guide mailing, included a supporting enhancement.
Another component of process is to monitor marketing activities and track progress toward enrollment and retention goals. The research team used Excel software (Microsoft, Redmond, WA, USA) to create recruitment-tracking spreadsheets that organized participants by group (experimental, control), geographic location (New Jersey, Arizona), and language (English, Spanish). Plotting spreadsheet data by group × location × language enabled researchers to assess progress, adjust allocation of recruitment resources (e.g., staff time, recruitment flyer printing and distribution, recruitment activities), and forecast when recruitment goals would be reached. Recruitment reports were generated and reviewed twice weekly to facilitate timely, informed research management decisions [
22]. In addition, these reports were scrutinized to cross-check data to detect and eliminate duplicate completions of the baseline survey from recruitment counts, a problem noted by others [
86,
106].
Creation of similar spreadsheets supported participant retention efforts. The website tracked participant use of the website, enabling researchers to download reports of activities that were organized in spreadsheets. Retention spreadsheets tracked each participant’s progress through the RCT and days elapsed since his or her last communication with project staff or website. Charting retention data permitted researchers to visualize participant progress throughout the time course of the RCT, monitor the rate at which participants completed each aspect of the RCT, and rapidly intervene when participants appeared to be experiencing barriers that may cause them to become inactive and drop out. The retention-tracking spreadsheet was annotated to indicate when and how inactive participants were contacted. For instance, when retention reports indicated a participant had not been to the study website for 30 days, staff contacted the participant via phone and email about every 10 days to encourage them to return to the website and continue participation. After 60 days of inactivity, staff reviewed the participant’s previous guide choices, mailed a new guide, and set the website to send the guide-specific nudges. This procedure was repeated until parents had completed all guides associated with a level, at which time staff contacted inactive parents by phone, email, and mail (using branded, bright blue, shiny Mylar envelopes [DuPont Teijin Films, Chester, VA, USA]) to encourage them to complete the next survey. As noted by others [
35], staff anecdotally observed that phone calls tended to result in participant action more so than emails. Staff were patient and persistent in their intensive efforts to retain participants, calling, emailing, texting, sending messages via Facebook® as needed to reach them and encourage continued participation [
23,
92,
107]. Retention reports were generated and reviewed twice weekly to facilitate timely contacts with participants as well as retention [
92] and reassessment of recruitment goals.