April 4-6, 2024 • Hyatt Regency • Lexington, KY
Innovations in Health Communication
Innovations in Health Communication
Abstract: AI in Health Communication: Prior Knowledge, Source Credibility, and Pro-Health Intentions
This research explores the perceptions and impact of visual health communication (VHC) initiatives incorporating artificial intelligence (AI) in motivating health behaviors, and the role of individuals' prior knowledge in this process. This project aims to investigate whether individuals with less prior health knowledge exhibit a more positive reception of AI-generated visual health communication messages compared to those with more prior health knowledge and whether the perceived source credibility of VHC influences recipients' attitudes.
This study seeks to answer these questions through a two-step approach: first a pre-test survey and then a randomized survey experiment. The pre-test will involve conducting a survey with the primary objectives of fine-tuning measurement and enhancing the assessment of outcomes. This survey will pre-test infographics generated using the generative AI program MidJourney; infographics are known for their effectiveness in health communication (Lee et al., 2021; Siricharoen & Siricharoen, 2017; Wansink & Robbins, 2016). The survey will be distributed through the Qualtrics platform and will ensure the quality of the measures and the clarity of the methods.
The larger project will be a randomized survey experiment designed to assess how revealing that images in VHC are AI-generated impacts participants' attitudes, perceptions, and intentions toward specific health behaviors. The survey will employ a 3 (topics: water consumption, exercise, vaping) x 4 (taglines for image source: AI, researchers, co-designed, none) between-subjects design, resulting in twelve distinct conditions featuring crossed messages and taglines. These topics were strategically chosen to align with prevalent health concerns in contemporary society (Schaeffer, 2019; Shriber, 2023; World Health Organization, 2022) making them ideal for the research objectives in the context of public health promotion and behavior change.
The taglines for the health messages will include the following options, in addition to a "none" condition:
"This message was created by a team of researchers."
"This message was generated by artificial intelligence."
"This message was a collaborative effort between researchers and artificial intelligence."
The inclusion of these tagline variations will provide the ability to comprehensively assess the impact of different message types on individuals' responses to AI-generated health communication initiatives.
This research holds significant implications for health communication, as AI continues to play an increasingly prominent role in shaping the way health information is disseminated (Kueper et al., 2020; Rai et al., 2023; Neves et al., 2018). By investigating the role of prior knowledge, this research aims to shed light on whether AI can effectively influence individuals with limited health information. Furthermore, the exploration of source credibility is essential for understanding how individuals perceive and trust AI-generated health messages.
The findings will contribute to the development of more effective AI-driven health communication strategies, potentially tailored to the knowledge levels of their intended audiences. This research has the potential to inform public health initiatives and campaigns, ensuring that health messages are not only accessible but also trusted by a wide range of recipients. In an era where health promotion is crucial, understanding how AI can optimize the impact of health communication is of paramount importance.
This study seeks to answer these questions through a two-step approach: first a pre-test survey and then a randomized survey experiment. The pre-test will involve conducting a survey with the primary objectives of fine-tuning measurement and enhancing the assessment of outcomes. This survey will pre-test infographics generated using the generative AI program MidJourney; infographics are known for their effectiveness in health communication (Lee et al., 2021; Siricharoen & Siricharoen, 2017; Wansink & Robbins, 2016). The survey will be distributed through the Qualtrics platform and will ensure the quality of the measures and the clarity of the methods.
The larger project will be a randomized survey experiment designed to assess how revealing that images in VHC are AI-generated impacts participants' attitudes, perceptions, and intentions toward specific health behaviors. The survey will employ a 3 (topics: water consumption, exercise, vaping) x 4 (taglines for image source: AI, researchers, co-designed, none) between-subjects design, resulting in twelve distinct conditions featuring crossed messages and taglines. These topics were strategically chosen to align with prevalent health concerns in contemporary society (Schaeffer, 2019; Shriber, 2023; World Health Organization, 2022) making them ideal for the research objectives in the context of public health promotion and behavior change.
The taglines for the health messages will include the following options, in addition to a "none" condition:
"This message was created by a team of researchers."
"This message was generated by artificial intelligence."
"This message was a collaborative effort between researchers and artificial intelligence."
The inclusion of these tagline variations will provide the ability to comprehensively assess the impact of different message types on individuals' responses to AI-generated health communication initiatives.
This research holds significant implications for health communication, as AI continues to play an increasingly prominent role in shaping the way health information is disseminated (Kueper et al., 2020; Rai et al., 2023; Neves et al., 2018). By investigating the role of prior knowledge, this research aims to shed light on whether AI can effectively influence individuals with limited health information. Furthermore, the exploration of source credibility is essential for understanding how individuals perceive and trust AI-generated health messages.
The findings will contribute to the development of more effective AI-driven health communication strategies, potentially tailored to the knowledge levels of their intended audiences. This research has the potential to inform public health initiatives and campaigns, ensuring that health messages are not only accessible but also trusted by a wide range of recipients. In an era where health promotion is crucial, understanding how AI can optimize the impact of health communication is of paramount importance.