The Risk Information and Skin-cancer Education for Undergraduate Prevention (RISE-UP) Study: Protocol for a Trial of Personalized Sun Protection Interventions for Skin Cancer Prevention among Undergraduate Students

Abstract

Background:

Addressing the increasing incidence of skin cancer among young adults is a priority. The objective of the Risk Information and Skin-cancer Education for Undergraduate Prevention (RISE-UP) study is to identify personalized intervention components to prevent sunburn, a clinically significant outcome highly associated with skin cancer, in college students.

Methods:

Guided by the Elaboration Likelihood Model, the study will use Multiphase Optimization Strategy (MOST) methodology to test three intervention components (ultraviolet photography, MC1R genetic testing, and action planning) each with two levels (yes v. no) in a full-factorial experiment to evaluate unique and combined effects of these components to improve outcomes over the longer-term, with seasonally timed follow-up. At-risk University of Utah students (N=528) will be recruited. Eligibility criteria include self-reported sunburn or tanning in the past year, or not utilizing recommended sun protection. After baseline assessment, participants will be randomized to intervention group, stratified by sex. Assessments will be completed at (1) Baseline; (2) Intervention; (3) 1 month after intervention; (4) 4 months after intervention (the end of the first summer); and (5) 15 months after intervention (the end of the second summer). The primary outcome will be participants’ self-reported number of sunburns. Secondary outcomes will include self-reported sun protection and tanning behaviors and, in a randomly selected subgroup, an objective measure of ultraviolet radiation (UVR) exposure.

Conclusion:

The RISE-UP study will determine the efficacy of different combinations of personalized skin cancer preventative interventions for young adults and determine the optimal combination of intervention components to prevent skin cancer.

Background

Skin cancer is the most common cancer in the US and worldwide and incidence is rising [1]. Basal and squamous cell non-melanoma skin cancers impart low risk of death but are expensive to treat, costing more than $8 billion in the US annually [2]. Non-melanoma skin cancers are increasingly common among young adults, and melanoma is increasing among young women [3, 4]. Melanoma, the deadliest form of skin cancer, is the 3^rd most common cancer among adults under age 30 [4, 5]. Ultraviolet radiation (UVR) exposure, particularly early in life, is the major risk factor for all skin cancers [6, 7]. Most melanomas are caused by UVR from sun exposure, but can also stem from artificial sources such as tanning beds [8].

Efficacious skin cancer prevention interventions for young adults are urgently needed. Although indoor tanning use is decreasing [9], young adults continue to be exposed to excessive UVR through poor use of sun protection and the pursuit of a tan during outdoor activities [10]. Young adults, especially men, frequently engage in outdoor unintentional tanning [11]. Excessive UVR exposure increases risk for sunburn, a clinically significant risk factor for skin cancer [12].

Many (40%-50%) US young adults are enrolled in college or graduate school, making university settings efficient contexts to reach young adults [13]. High levels of risk behaviors take place in college settings [14] and the Surgeon General identified college students as needing evidence-based skin cancer prevention interventions [15]. College students often have a culture of tanning and high levels of participation in outdoor social, recreational, sporting, and employment activities [16]. Targeting young adults in college captures them at a critical developmental stage where life-long habits are being consolidated and desires for highly personalized content peak [16, 17]. College students are also highly interested in testing new health technologies, including those incorporating genetics [18].

There is a documented benefit of behavioral counseling for skin cancer risk reduction in this age group [19]. However, prior skin cancer preventive interventions have relied on a one-size-fits-all approach with multiple intervention components delivered simultaneously. Interventions for young adults should ideally be highly personalized to be consistent with their expectations and receptivity for such content [16, 17]; however, existing interventions have not yet tested different combinations of personalized content [17, 20]. Guided by the Elaboration Likelihood Model (ELM) [21], the underlying conceptual premise of the current study is that personalization of risk information will best galvanize skin cancer prevention behaviors in young adults via central processing marked by cognitive deliberation and elaboration of personalized information (Figure 1).

Figure 1.

Conceptual Model

Rationale for Selection of Candidate Intervention Components

There are three highly personalized intervention strategies that hold the most promise for young adults: UV photography that highlights skin damage and the individual’s skin cancer risk, provision of genetic risk feedback related to variants in the melanocortin-1 receptor (MC1R) gene, and personal action planning [22-25]. Personalized UV photographs have been used with college students to alert them to existing UVR-induced damage in their skin. While UV photographs have led to improved skin cancer prevention behaviors, effect sizes have been small [22, 26]. However, most prior studies have emphasized the potential for skin damage to worsen over time rather than tie the damage to skin cancer risk [22], as we do in the current study.

Meta-analytic results indicate that provision of personalized genetic risk feedback motivates health behavior change [27], but there is little data on young adults. Information on inherited variation in the MC1R gene is a novel risk tool that could increase sun protection and tanning avoidance in young adults. Individuals with variants in MC1R have a higher risk for developing skin cancer [28]. Individuals are interested in MC1R testing, motivated to use it to make skin health decisions [29], and the results lead to increased sun protection use and skin screening, and reductions in sunburn [23-25].

Interventions to engage implementation intentions (action planning) involve the creation of personalized plans for implementing health behaviors, including what behaviors an individual will engage in and in what contexts they will do so. Action plans have been used to improve health behaviors among college students (e.g., alcohol/marijuana use) [30]. However, action plans have not yet been used with college students to enhance planning to increase sun protection and avoid tanning.

In the current study, a factorial experiment design drawing on the Multiphase Optimization Strategy (MOST) framework [31], will be used to test which of the three interventions described above or their combination have the most impact on desired outcomes. We selected a factorial experiment because it improves intervention efficiency [31] and requires substantially fewer subjects to achieve similar power to test component effects compared to designs such as RCTs. The inclusion of three personalized components is guided by conceptual processes inherent to the ELM [21], which highlight the persuasive impact of central processing of personal risk information as well as enhanced efficacy and control regarding cancer risk reduction behaviors (see Figure 1). The intervention components also target distinct and complimentary aspects of risk – environmental risk due to past UVR exposure (UV photo), genetic risk based on inherited skin cancer susceptibility (MC1R genetic testing), and risk due to behavioral habits (action plan). In our prior pilot work, we found that both UV photo and MC1R testing were associated with improvements in sun protection and/or tanning, but that a combined approach of both showed particular promise [32]. However, it is unknown which of these interventions plus action planning or their combinations may best prevent sunburn.

Goals of the Current Study

The primary study goals are to 1) examine the effect of personalized risk components on college students’ sunburn to determine which intervention components (UV photo, MC1R genetic testing, action plan) lead to sunburn prevention and reduction of UVR exposure, including through a follow-up after the high UVR summer one-year later, 2) examine how these intervention components influence other important skin cancer prevention outcomes, including reduction of tanning behaviors, improved use of sun protection, and UVR exposure measured with objective monitoring, and 3) examine moderators (e.g., biological sex) and mediators (e.g., central processing) of intervention efficacy.

Methods

Study Design

A full factorial experiment with $2^3=8$ experimental conditions (Table 1) will be used to test which of 3 intervention components or their combination eliminate sunburn over one year and impact tanning, sun protection, and objectively-measured UVR exposure. All participants will also receive a core intervention consisting of educational materials (see below) and be randomly assigned to each component level (yes v. no).

Table 1.

Intervention Conditions

Arm	Education	Action Plan	UV Photo	MC1R Test
1	Yes	Yes	Yes	Yes
2	Yes	Yes	Yes	No
3	Yes	Yes	No	Yes
4	Yes	Yes	No	No
5	Yes	No	Yes	Yes
6	Yes	No	Yes	No
7	Yes	No	No	Yes
8	Yes	No	No	No

Participants and Setting

Participants will include undergraduate students enrolled at the University of Utah (UofU). Demographics of UofU students are similar to other public US universities (Table 3) [33]. UofU students will be eligible to participate if they are: (1) at least 18 years old, (2) are enrolled as an undergraduate student, and (3) report skin cancer-related risk behavior based on current skin cancer prevention recommendations [34, 35], including having at least one sunburn in the last year, and/or reporting indoor tanning at least once in the last year, and/or reporting intentional or unintentional outdoor tanning “sometimes,” “often,” or “always,” and/or reporting using sunscreen plus one or more other sun protection behavior (protective clothing, shade) infrequently (“never,” “seldom,” or “sometimes”). Students will be excluded if they have a personal history of skin cancer or do not read or speak English.

Table 3.

Demographic characteristics of University of Utah undergraduate students

Characteristic		N (%)
Gender	Female	10,769 (47)
Race/Ethnicity	Non-Hispanic White	15,637 (68)
	Other	7,448 (32)
Median Age	Full-time and part-time students	21

Utah has the highest melanoma rate in the US [36] and one of the steepest increases in melanoma incidence, including among those under age 30 [5, 37]. This is due to high levels of UVR exposure due to geographic elevation and availability of outdoor activities year-round. The incidence of melanoma and abundant opportunities to use sun protection and decrease tanning make this region an optimal setting to test melanoma preventive interventions.

Procedures

Study recruitment will occur January-March and will be advertised through flyers, classes, and university organizations. Students who screen eligible will be asked to schedule an in-person baseline visit (T1), during which they will provide informed consent and complete a questionnaire in REDCap [38]. Participants will be assigned to one of 8 intervention arms (stratified by sex-assigned at birth; equal assignment to all arms) using a randomization tool in REDCap (Table 1). Participants randomized to receive UV photo or MC1R testing will have their photo taken using the UV camera and/or provide a saliva sample via a Genotek Oragene kit, respectively. Follow-up assessments will occur immediately after the intervention (T2) during the videoconference visit, and at 1 month (T3), 4 months (T4), and 15 months (T5) post-intervention (Figure 2). T3, T4, and T5 assessments will be sent to participants electronically (email, text). Assessment completion reminders will be sent after 3, 6, and 7 days. This study and its procedures were approved by the IRB at UofU and the relying IRB at MSKCC.

Figure 2.

Study Procedures

Intervention

Intervention visits will be conducted by a bachelor’s level research assistant via videoconference (e.g., Zoom) in April and May.

Core Intervention

Participants will receive education on skin cancer and prevention, including use of sun protection strategies and tanning avoidance, based on American Academy of Dermatology and CDC resources [39, 40].

Candidate Intervention Components

UV Photo.

Participants assigned to the “yes” level of this component will view a photo of their face in visible light and UV light (Figure 3) from the VISIA Complexion Analysis system. The presentation will provide an explanation of UV light and harm to the skin, that the VISIA system reveals damage that is not typically visible, that darker areas indicate where UV damage has occurred, and that using sun safe behaviors early in life will have greatest impact and decrease skin cancer risk.

Figure 3:

Intervention Materials

MC1R Genetic Testing.

Prior to the intervention visit, saliva samples for all participants assigned to the “yes” level of this component will be analyzed by the Diagnostic Molecular Genetics Laboratory of Memorial Sloan Kettering Cancer Center (MSK). Complete sequencing of MC1R coding region (1 exon, 961 bp) will be performed to identify all existing variants. Based on meta-analytic literature, higher risk MC1R variants are: p.V60L, p.D84E, p.V92M, p.R142H, p.R151C, p.I155T, p.D160W, p.R163Q, and p.D294H. Participants’ MC1R results will be classified as “higher genetic risk” (those who carry at least one higher risk MC1R variant) or as “average genetic risk” (those who have other MC1R genotypes)[41, 42]. The higher risk MC1R variants confer a 1.4 to 2.4-fold higher risk for melanoma and other skin cancers compared to other genotypes, and an average absolute risk of 3% compared to 1% [43]. Consistent with risk feedback strategies we and others have established and tested [44, 45], MC1R genetic test result findings will be provided including the general risk category (higher or average) in lay language in a presentation and handout that addresses personal risk level and population risk level (Figure 3) and strategies to reduce skin cancer risk. Participants who report distress related to their result will receive contact information for clinical genetic resources.

Action Plan.

Participants assigned to the “yes” level of this component will receive guidance from the research assistant in completing worksheets for personalized action plans for sun protection and tanning. For sun protection (Figure 3), participants select an outdoor activity, which sun protection behaviors they plan to use, when and where they will use them, and what supplies they need for the sun protection. For tanning, participants identify tanning locations and activities, select a tanning behavior to work on, and plan for an alternate activity that meets their goals while minimizing tanning and using sun protection. Those who report not tanning will not complete a tanning action plan.

Core Measures (Table 3)

Outcome measures.

Sunburn is the primary outcome. Participants will report on the number of sunburns they experienced (0, 1, 2, 3, 4, 5 or more) in the past month using an item from the Sun Habits Survey [46], a well-established and valid measure [47, 48]. Frequency of sunburn in the past year will be assessed using an item from the NHIS 2015 survey [49] with response options of 0, 1, 2, 3, 4 or more.

Sun protection behaviors, objective UVR exposure, and tanning are secondary outcomes.

Sun protection behaviors and sun exposure habits in the past month will be assessed using a modified version of the Sun Habits Survey [46]. Participants will be asked to report on sun protection behaviors (sunscreen, re-application of sunscreen, shirt with sleeves that covers the shoulders, long pants or a long skirt, hat, shade, avoiding being in the sun between 10:00am–4:00pm, sunglasses) used when they were in the sun for 15 minutes or more using a 5-point Likert-type scale (“never” to “always”). Items will be summed to create a sun protection composite. Participants will be asked to report on time spent outdoors on weekends and weekdays. Items were altered from the Sun Habits Survey [46] to assess time spent outdoors in the past month.

Objective UVR assessment will be collected from a random sample of 25% of participants stratified across study arms who will be assigned to wear a UV dosimeter for 7 days immediately after T3 [50]. Participants will be asked to wear the device during waking hours and complete a daily evening survey assessing device-wearing during outdoor time, sun protection measures used, and sunburn.

Tanning behaviors, including intentional indoor and outdoor tanning, will be assessed using modified items from a well-established indoor tanning measure and the Sun Habits Survey [46, 51]. Participants will be asked to report on past month intentional outdoor tanning on a 5-point Likert-type scale (“never” to “always”), and intentional indoor tanning with response options of “0 times,” “1 or 2 times,” “3 to 9 times,” “10 to 19 times,” “20 to 39 times,” or “40 or more times.” Additionally, we will assess indoor tanning frequency in the past year using an item from HINTS with response options of “0 times,” “1 to 2 times,” “3 to 10 times,” “11 to 24 times,” or “25 or more times.” We will assess unintentional outdoor tanning (i.e., ending up with a tan when not intending to get tan) in the past month using modified items from a previous study that examined college tanning behaviors with responses on a 5-point Likert-type scale (“never” to “always”).

Sociodemographic factors

We will collect participant demographics including sex-assigned-at-birth, gender, race, and ethnicity during screening. All other demographics will be collected on the baseline questionnaire and will include socio-economic characteristics (e.g., marital status, family income, employment status), personal history of cancer, family history of skin cancer, prior experience with genetic testing, baseline patterns of risk behavior (intentional vs. unintentional tanners), and the Fitzpatrick scale [52]. We will use demographics to describe the sample and as potential effect moderators.

Putative mediators

Central processing.

Central processing, consistent with ELM-driven elaboration and consideration of personal risk information [21, 53] will be assessed through a series of items addressing personal risk deliberation. These include frequency of discussions with family and friends regarding risk feedback on a 4-point Likert-type scale from “a lot” to “not at all” [54, 55] and the content of these discussions (e.g., skin cancer risk reduction topics covered) [56]. Participants will also be asked how much they agree with six statements on their comprehension and trust in study materials (e.g., “It took a lot of effort to understand the RISE-UP materials provided today on skin cancer and prevention”) [54, 55]. Participants will be asked how much they agree with 11 items that measure cognitive processing of risk information (e.g., “Knowing about my chances of getting skin cancer is not important to me”) [54, 57, 58]. We also will assess frequency of thought about study feedback assessed on a 7 point scale from "not at all" to “all of the time” [56]. We also will assess frequency of thought about study feedback assessed on a 7 point scale from "Not at all" to “All of the time” [56].

Additional constructs related to central processing include perceived risk, self-efficacy, and perceived control. Perceived risk will be assessed using 4 items. One item will assess absolute risk on a 7-point Likert-type scale (“No chance” to “Certain to happen” with option of “Don’t know”) and another item will assess comparative risk on a 5-point Likert-type scale (“My chance is well below average” to “My chance is well above average” with option of “Don’t Know”). Three items will assess affective risk (e.g., “I feel very vulnerable to developing skin cancer someday”) [59], Self-efficacy related to sun protection use and tanning avoidance will be assessed. Participants will rate their confidence completing sun protection behaviors (e.g., “Use sunscreen even if you do not like how it feels”, “Cover up with protective clothing even when it is hot outside”) on a 5-point Likert-type scale (“not at all confident” to “extremely confident”).” [60-62]. Perceived control over development of skin cancer will be assessed using 3 items (e.g., “There’s not much you can do to lower your chances of getting skin cancer) on a 5-point Likert scale from “strongly disagree” to “strongly agree” [63].

Knowledge will be assessed using 17 true/false items created by the investigators based on the study’s skin cancer prevention educational materials [39]. Peripheral processing will be assessed using 6 items from previous studies that evaluate general interest in and satisfaction with study materials (e.g., “The RISE-UP materials are/were interesting”, “I find/found the RISE-UP materials easy to understand”) on a 7 point Likert-type scale (“very strongly disagree” to “very strongly agree”) [64, 65].

Exit interview.

Participants who complete the study will be invited to participate in an exit interview on topics such as what participants learned from the study, how they felt about the information, how much they applied it to their lives, and feedback on study procedures.

Statistical Approach

Power and sample size

Sample size was selected to ensure power >80% for detecting a risk difference for the primary and clinically significant outcome of sunburn from baseline to month 12 of 0.15 or greater (e.g. 55% vs. 40% rate of sunburn from baseline to month 4) between intervention component “present” and “absent” experimental conditions at a type I error rate of 0.05. We assume a rate of attrition at 4 months of <25%, consistent with our experience and other studies in the college student population,[66, 67] so that 528 total students randomized (66 students randomized to each arm of the factorial design) will ensure >352 total participating students completing 4 month follow-up (>44 students completing 4 month follow-up per factorial arm). Randomization will be stratified by sex-assigned-at-birth to ensure balance between groups and to maximize power for sex-assigned-at-birth moderation analyses. The proposed study will also be well-powered to detect two-way interactions between candidate intervention components and effect modification by sex-assigned-at-birth. In particular, the proposed study will provide >80% power for detecting risk difference moderation or two-way interactions of 0.22 or greater at month 4 (e.g. 62% vs. 40% rate of sunburn for intervention component among women as compared to 50% vs. 50% rate of sunburn for intervention component among men).

Analytic approach.

The primary analyses will be performed from an intention-to-treat perspective, where each participant will be analyzed according to assigned study component(s). Intervention component main effects will be assessed in the context of an ordinal logistic or logistic regression model (depending on the distribution of number of sunburns observed) with indicators for each of the three candidate intervention effects and adjusted for sex. The primary analyses will include a main effect for ambient UVR, which will be measured at each time point by collecting current postal zip codes from participants during their assessments. The zip codes and the date that zip codes were provided will be used to analyze UV index in the area on the date the participants complete their assessments. The proposed study is intended to screen for promising intervention components to optimally prevent sunburn, a clinically significant outcome related to the development of skin cancers, and as such will make no adjustment for family-wise type I error control due to multiple testing. In this study, a type II error (mistakenly failing to identify an active intervention component) is as important as a type I error. The optimized candidate intervention identified in this study will be validated in a follow-up effectiveness study.

Two- and three-way interactions between intervention components and effect moderation of intervention components by sex will be assessed using analogous logistic models that include the relevant interaction term (e.g., sex by UV photo or Action Plan by MC1R genetic testing). The primary endpoint will be the T3 assessment, and the analyses will account for baseline sunburn levels. We will explore other moderators including baseline tanning behavior and sun protection, as well as skin type. Analogous analyses will be conducted for secondary endpoints including changes from baseline to T4 and baseline to T5 in tanning behaviors, sun protection, and objective UVR exposure, as well as sunburn at 4 months post-intervention, in the context of linear models. Similarly, pre-planned subgroup analyses for the primary, clinically significant endpoint (sunburn) and secondary endpoints (changes in tanning behaviors, sun protection, objective UVR exposure) will be conducted within sex and baseline tanning behavior (intentional vs. unintentional tanners).

Assessments of mediators of intervention effects on primary and secondary endpoints will be assessed via causal mediation analyses [68]. The direct and indirect effects of intervention components on outcomes (e.g., sunburn) that are mediated by elements of central processing will be estimated to evaluate the contribution of each path to the overall effect of each intervention component on outcomes [69, 70]. Analyses will examine the potential effect that wearing a UV dosimeter may have on outcomes. Since this approach will also require a statistically more stringent assumption of sequential ignorability (i.e., no unmeasured confounders), which stipulates that variables included in the models must be sufficient to control for extraneous factors that jointly influence mediators and outcome, sensitivity analysis will be performed to evaluate the effects of violation of this assumption. The primary mediation analysis will focus on T3, with secondary analyses examining mediation at T4 and changes in mediators from T1 to T3 and changes from baseline T1 to T4.

Discussion

Young adults are at elevated risk for skin cancers [10, 14]. The current study will test the efficacy of education on skin cancer prevention plus the individual and combined effects of personalized intervention components among young adults in the college setting to provide the basis for an optimized skin cancer prevention intervention that is efficacious and scalable. Another study contribution is the examination of intervention moderators and mediators to inform future intervention tailoring. We include robust indicators of ELM central and peripheral processing of risk feedback information to carefully assess mediational processes by which optimized intervention outcomes occur. In this way we will have generalizable information about how best to “drive the message home” in future skin cancer prevention studies. In terms of mediators, prior work has examined increased tanning knowledge and perceived susceptibility to skin damage [22, 71, 72].

Study strengths include the focus on at-risk college students where behavior change is important to their long-term health. In addition, this study uses a MOST design, which employs highly efficient factorial experiments to systematically evaluate intervention components’ individual and combined effects [31]. Another strength is the focus on sunburn occurrence as a primary outcome, given that sunburn is the primary modifiable factor associated with skin cancer and melanoma risk [12, 73]. Our study assesses outcomes over 15 months to establish intervention sustainability; the vast majority of prior studies have not included follow-up periods as long as a year or during high-risk times (e.g., summer, when opportunities for sunburn increase). And finally, this study employs an objective assessment of UVR exposure in a subsample of participants, which will complement self-reported exposure information by providing multi-method data. Information from the monitors could also inform future interventions for skin cancer prevention in this population, by elucidating patterns of UVR exposure among college students and identifying “higher risk” days and times when students receive the highest UVR exposure and could benefit from targeted interventions.

Study limitations include a reliance on self-report for the primary outcomes. While self-report measures, such as reported sunburn occurrence, have been shown to be valid indicators of exposure [47, 74], we also employ a random sample of participants (due to feasibility and cost constraints) stratified across study arm to wear a UVR monitoring device [50]. Additionally, the study focuses exclusively on college students, a sizable proportion (40%-50%) but still a subset of the young adult population [13, 75]. Study participants may have a variety of baseline behavioral characteristics that increase their risk for skin cancer (e.g., sunburn, tanning, poor use of sun protection), and if so, we will examine whether these baseline behaviors moderate study outcomes. Additionally, multiple study arms may yield benefits to the primary outcome of sunburn occurrence and other outcomes. If this is the case, the best combination of intervention components for further examination will be determined based on factors such as intervention parsimony, feasibility and cost of intervention components, and qualitative feedback from participants.

The RISE-UP study will be the first to apply the MOST framework to test the relative impact of personalized skin cancer risk intervention components among young adults. Based on study findings, the identified optimal set of RISE-UP intervention component(s) will next be tested in a fully-powered RCT in higher risk young adults. We plan to extend this line of research to other regions of the US as well as to young adults who are in non-university settings. The moderator findings will enable us to adapt RISE-UP for use in important subpopulations that have the most to gain and to target individuals who require additional intervention. Future RCTs could include dissemination and implementation assessments to understand multi-level barriers and facilitators of use of RISE-UP across diverse settings where young adults may be engaged, emphasizing a health equity implementation focus. Finally, future work could develop automated methods for communicating personalized risk information and interventions to eliminate the need for in person visits and to increase dissemination.

Table 2.

Core Study Measures

Core Measures	Timepoints for Assessment
	T1	T2	T3	T4	T5
	Primary Outcomes (Aim 1)
Sunburn occurrence		✓	✓	✓	✓
	Secondary Outcomes (Aim 2)
Sun protection behaviors and sun exposure habits		✓	✓	✓	✓
Tanning behaviors		✓	✓	✓	✓
	Moderators (Aim 3)
Sociodemographic factors (e.g. sex), baseline tanning & sun protection, skin type
	Mediators (Aim 3)
Perceived risk deliberation		✓	✓	✓	✓
Perceived risk		✓	✓	✓	✓
Self-efficacy		✓	✓	✓	✓
Perceived control		✓	✓	✓	✓
Knowledge		✓	✓	✓	✓
Peripheral processing		✓	✓	✓	✓