Image generated with Dreamstudio. Prompt: "A dazzling explosion of color and brightness to be used as a Web banner."

When conditioning is not needed, stimuli can be built even faster! 

This is a Web version of the Companion Appendix for RAISE: A New Method to Develop Experimental Stimuli for Advertising Research with Image Generative Artificial Intelligence, developed by César Zamudio, Jamie Grigsby, and Meg Michelsen, and published in the Journal of Advertising. Questions? Feel free to contact us!
​Before discussing further examples of experimental stimuli developed using RAISE, it must be emphasized that conditioning on the image of Stimulus A is not always required. For instance, consider the work of Cowart (2020), who studies the role of paternal referents in advertising, manipulating paternal referents as a mom and a dad. While, in the original RAISE example of waiters, our methodology ensured that the waiters created across conditions were very similar, strict control of the features of the visual stimuli might not be necessary.

​If strict control across conditions is not imperative, then Stimulus A can be created first through an initial prompt:
Picture
Stimulus A prompt: The picture of the face of a suburban mom, with short hair, about 37 years old, very happy, with a wide smile. She is white, and is wearing a blouse. She is brunette. Her face is shown. White background, transparent background, blurred background.
​Afterwards, Stimulus B can be created sequentially, and conditionally, but only on the initial text prompt of Stimulus A, without conditioning on the image of Stimulus A:
Picture
Stimulus B prompt: The picture of the face of a suburban dad, with short crewcut hair, about 37 years old, very happy, with a wide smile. He is white, and is wearing a white shirt. He is brunette. His face is shown. White background, transparent background, blurred background
The pictures above show a replication of the mom and dad picture stimuli used by Cowart (2020) – as can be seen, only the initial text prompt of Stimulus A, with slight variations, is sufficient to construct Stimulus B. Below is a handy summary card outlining the process you just learned:
Picture
Summary card to create two different persons: a mom and a dad, to be used in different ads. Feel free to save and share!
These two stimuli can then be used in conjunction with applications such as PowerPoint or Adobe Illustrator to fully construct experimental stimuli (in the case of Cowart (2020), microwave oven ads: here are the final compositions replicating the authors' work, which were constructed using similar assets and PowerPoint, and can be used in experiments or A/B testing:

Ready to learn more?

Feel free to browse our gallery of examples with full tutorials, and the technical details section to learn the finer points of generating visual experimental stimuli using the RAISE methodology.