Background
Body image in Brazil
Digital innovation in body image interventions
The current trial design
-
H1: The chatbot was designed to provide immediate benefits to users. Therefore, it is anticipated that adolescents will experience improvements in state-based body satisfaction and affect at the time of engagement with the chatbot.
-
H2: Adolescents randomised into the intervention condition will experience greater improvements in trait-based body esteem, affect and body image self-efficacy at post-intervention, one-week and one-month follow-up, relative to the assessment only control condition.
-
H3: Based on previous research into the moderating effects of gender [30] and trait psychopathology [31, 32] on body image intervention effectiveness, it was hypothesized that intervention effects will be moderated by gender, and baseline levels of body esteem, affect and body image self-efficacy. Specifically, intervention effects will be greatest among: girls; girls and boys with lower levels of trait body esteem; girls and boys with higher levels of trait negative affect and body image self-efficacy.
-
H4: With regards to user engagement and adherence, given the novelty of this intervention, analyses will be of an exploratory nature. It is, however, anticipated that greater engagement and adherence with the chatbot intervention will lead to greater improvements in trait-based outcomes.
Method
Public and patient involvement
Eligibility criteria
Intervention
Intervention development and content
Phase | Key development tasks and decision-making processes |
---|---|
1. Select and appraise the best available scientific research and evidence. | • Critically evaluated extant literature to identify key risk and protective factors for body image among Brazilian adolescents. • Concurrently, UNICEF conducted 26 workshops with 681 (66% girls) ethnically diverse Brazilian adolescents across 15 northern and south-eastern cities to better understand the nature of, and influences on, their body image, and their preferences for the chatbot’s key messages and persona [34]. |
2. Apply the evidence in a collaborative decision-making process with relevant parties. | • Data from the literature review and focus groups informed a collaborative decision-making process between academics, industry partners and UNICEF to select which risk and protective factors would be targeted within the chatbot. • Three themes were agreed upon: 1) Unrealistic beauty standards perpetuated by society; 2) Appearance pressures experienced within interpersonal relationships; 3) Acceptance and appreciation of one’s body. |
3. Selection and adaptation of extant techniques that target the agreed upon themes, for use in the chatbot. | • A micro-intervention framework was used to identify and select therapeutic techniques that addressed the above themes [35]. • Techniques needed to be brief (e.g., completed in under 10 min), standalone (e.g. a distinct beginning and end), immediately actionable (e.g., can be executed in the moment, without additional resources) and adaptable for digital platforms. • Researchers drew on their expertise of working with evidence-based body image prevention and intervention approaches and a meta-analysis on therapeutic change techniques to identify and select appropriate techniques [31]. • Eight techniques (see Table 2) were extracted from extant evidence-based body image programs, including: Confident Me [36], The Body Image Workbook [37], and Expand Your Horizon [38]. • A script was developed for each technique and was reviewed by body image experts from Brazil and the UK, as well as Dove and UNICEF. • Scripts were then amalgamated into a written interactive dialogue to be delivered by the chatbot fictional hosts, Dandara (a young Brazilian woman) and Gabriel (a young Brazilian man). • Hosts’ gender and persona were informed by: ° Mixed-gender preferences observed among young people receiving body image interventions [39]. ° Adolescents’ use of peers for emotional and informational support in online environments ([40]). ° Qualitative and quantitative data obtained by UNICEF via focus groups and an online survey via UReport with girls and boys name preferences. |
4. Development of the digital interface | • The digital interface of the chatbot was constructed by communication and technology experts at Talk2U. • This iterative process involved frequent reviews by adolescents, body image experts based in Brazil and the UK, Dove and UNICEF. • Key design features included: ° The use of Facebook Messenger, an instant messaging app for smartphones, to host the chatbot. ° The use of interactive elements that enhance user engagement and saliency of key therapeutic messages, including check-in messages, multi-media stimuli (i.e., audio and video clips; emojis [a small digital image or icon used to express an idea or emotion]), virtual hosts and the gamification of intervention techniques. ° In regards to gamification (i.e., the use of game design elements in non-game contexts, [41]), users needed to successfully complete a technique before progressing to the next technique within a thematic cluster. Gamification is used to incentivise users to continue their engagement while allowing them to master easier skills before progressing to more challenging tasks [42]. The ordering of techniques within each thematic cluster was informed by the researchers’ expertise, as well as feedback from further focus groups with adolescents on the acceptability and feasibility of the techniques and presentation formats. • Project partners with technology and marketing expertise advised on the selection of the chatbot’s name, Topity. • To reduce bias the institutional and industry affiliations were not displayed on the user interface. Participants and parents were, however, informed of the multi-international partnership and the relevant organisations involved in the ‘Parent and Participant Information & Consent’ form. • Topity can be accessed via Facebook Messenger using the following URL link: https://www.facebook.com/topitychat/ |
Topics | Change Technique Objectives | Technique Steps |
---|---|---|
Family, friends & body image | ||
1. Banish body talk | 1. Define and identify forms of body talk. 2. Identify the consequences of body talk and implement strategies for challenging this dialogue. | • User reflects on their own body talk, by estimating the frequency of engagement (e.g., number of times in preceding week), areas of fixation (e.g., belly; legs; skin colour) and the type of commentary (e.g., compliment, criticism, comparison). • Bot led deconstruction task (social media post); user considers the use and frequency of body talk and its impact on the person posting and their followers. • User generates alternative comments that focus on non-appearance aspects in the image (e.g., their affect; the activity, the location); user commits to engaging in this behaviour over the coming week. |
2. Dealing with provocative people | 1. Identify people that negatively impact your body image. 2. Explore how and when to use assertiveness strategies to address the unhelpful behaviours of others. | • Bot led discussion on how others impact our body image and how these behaviours are framed (e.g., concern, helpful advice, friendly teasing). • User identifies those who negatively impact their body image. • Bot introduces and applies the RIGHTS assertiveness model to an example scenario. • User invited to practice the RIGHTS model with the bot. |
Social media & body image | ||
1. The messages behind the media | 1. Develop media literacy skills in critically analysing and evaluating media content and the motivations of content creators (e.g., influencers, industry brands). 2. Identify and implement strategies that challenge unrealistic media images and messages, and create safe social media environments. | • Bot led discussion on media literacy; user identifies how media images are manipulated and the motivations of content creators. • Dove Selfie; user watches a 75-second clip on photo editing strategies, indicates their emotive reaction and identifies newly learnt strategies. • Bot describes strategies for creating a realistic, diverse and inclusive social media environment; user commits to engaging in one behaviour in the next week. |
2. Unfair-to-compare | 1. Define and identify examples of the cognitive distortion, ‘Unfair-to-Compare’. 2. Implement strategies that challenge and reduce upward comparison making. | • Bot led discussion on the internalisation of appearance ideals and upward comparison making. • User reflects on their engagement with this thinking style and the consequences associated with upward comparison making. • Bot describes strategies for challenging the ‘Unfair-to-Compare’ thinking style. |
3. A whirlpool of comparisons | 1. Identify thoughts and behaviours that contribute to the cycle of comparison making. 2. Identify and implement strategies to stop the cycle from starting or worsening. | • Bot led discussion on the cycle of comparison making; whirlpool analogy. • User identifies their comparison making behaviours and the consequences of these behaviours for themselves and others. • Bot describes strategies for exiting the whirlpool. |
4. The magnifying glass | 1. Define and identify examples of the cognitive distortion, ‘The Magnifying Glass’. 2. Identify and implement strategies that reduce selective attention on disliked body parts, while increasing attention towards areas that are liked and appreciated. | • Bot led explanation of selective attention and its role in maintaining body image concerns. • User reflects on their areas of fixation (e.g., stomach, legs, body hair). • Bot describes strategies for reducing selective attention. • User invited to engage in a guided mirror exposure task; if unable to in the moment, the audio is saved for a more appropriate time. |
Body appreciation and functionality | ||
1. Beauty bound | 1. Define and identify examples of the cognitive distortion, ‘Beauty Bound’. 2. Identify and implement strategies that reduce self-imposed body image rules that prevent life engagement. | • Bot led discussion on self-imposed body image rules and limitations; the importance of focusing on the body’s functionality (i.e., what it can do and experience) when attempting to overcome body image rules. • Visualisation activity; user recalls personal example of Beauty Bound thinking (e.g., thoughts, feelings, behaviours and outcome); then visualises the scenario without self-imposed boundaries (e.g., thoughts, feelings, behaviours and outcome); user identifies the differences. |
2. Expand your horizon | 1. Define body functionality and identify the different forms of functionality. 2. Describe the importance of our own body’s functions. | • Bot led discussion on body functionality (i.e., what the body can do and experience). • Functionality activity; user provided with categories of functionalities and the associated experiences / activities: sensation (i.e., touch, taste, sight); physical activity (e.g., running, cycling, jump-rope), creativity (e.g., cooking, photography, drawing) and self-care (e.g., petting animals, meditation, having a bath). • User selects their most important experience/ activity within each category. • User then reflects on one experience / activity in-depth (e.g., describe what dancing allows the person to experience and why it’s important to them). |
Intervention pilot study
Measures
Demographics | Age, gender identity, ethnicity, race, state and region of residence | T1 |
---|---|---|
Primary outcomes | ||
State body satisfaction | A single 11-point scale assessing the immediate impact of a chatbot technique on participants’ state body satisfaction (How satisfied are you with your appearance, right now, in this moment?). Total score range 0 (not at all) to 10 (very much). Higher scores reflect greater satisfaction. | Start of chatbot experience. Post-intervention technique. 12-, 16-, 23.5-h after non-engagement. |
Trait body satisfaction | Appearance Positive (AP; e.g., I like how I look like in photos), Appearance Negative (AN; e.g., I feel ashamed about my appearance) and Weight (e.g., I am satisfied with my weight). Mean subscale scores range between 0 (never) and 5 (always). Higher scores reflect higher esteem. Internal consistency: AP = .85; AN = .88; Weight = .89. | T1- T4 |
Secondary outcomes | ||
State affect | A single 11-point scale assessing the immediate impact of a chatbot technique on participants’ state affect (How happy are you, right now, in this moment?). Total score range between 0 (not at all) to 10 (very much). Higher scores reflect greater positive affect. | Start of chatbot experience. Post intervention technique. 12-, 16-, 23.5-h after non-engagement. |
Trait affect | The Positive and Negative Affect Scale for Children 8-item; PANAS-C8 [48] 8-items related to emotive states (4 positive [e.g., joyful]; 4 negative [e.g., irritated]). Positive and negative affect subscale scores range between 1 (not at all) and 5 (extremely). Higher scores reflect greater positive or negative affect. Internal consistency: PA = .77; NA = .76 | T1 - T4 |
Trait body image self- efficacy | A purpose-built measure, The Body Image Self-Efficacy Scale, derived from existing self-efficacy measures [49‐51]. 5 visual analogue scales (VAS; 0 [not at all confident] to 100 [very confident]) assessing participants’ belief in their ability to execute strategies to improve their body image (e.g., I am able to learn and practice new skills to help improve my body image). Total mean scores range between 0 and 100. Higher scores reflect greater body image self-efficacy. Internal consistency: .82 | T1 - T4 |
Intervention adherence | Digital metrics will assess participants engagement with the chatbot, including: • Percentage of participants who complete the intervention (e.g., minimum of 1 technique completed over the 72-h intervention period); • Percentage of participants who enter the intervention but do not complete the intervention; • Percentage of participants who complete each intervention technique; • Average number of techniques completed; • Average length of time taken to complete each technique; • Average length of time spent engaging with the chatbot over the 72-h period. | 72-Hour Intervention Period |
Intervention acceptability | 11 items assessed participants acceptability of the chatbot. Factors included: • Emotive response to the chatbot (e.g., enjoyment, interest, comfortability); • Relevance (e.g., pertinence, importance and helpfulness of the content to the user and other young people in Brazil); • Ease of use (e.g., speed and accuracy of responses); and • Willingness to recommend (e.g., how likely the user would recommend Topity to a friend and re-engage in the future). Item scores range between 1 (e.g., not at all enjoyable) and 7 (e.g., very enjoyable). Higher scores reflect greater acceptability. | T4 |