广告
图像(数学)
生成语法
心理学
生成模型
计算机科学
人工智能
业务
作者
César Zamudio,Jamie L. Grigsby,Meg Michelsen
标识
DOI:10.1080/00913367.2024.2438002
摘要
Advertising research widely uses visual stimuli. Stimuli development, whether by researchers or by hired designers, requires considerable time, funding, and know-how. Image generative artificial intelligence (iGenAI) allows faster and more cost-effective stimuli production, but whether this technology can produce rigorous experimental stimuli comparable to researcher-generated stimuli remains an open question addressed herein. First, we review publications in three advertising and marketing journals to identify relevant domains where iGenAI can be applied. Second, we present RAISE (Rapid Artificial Intelligence Stimuli for Experiments), a new methodology to generate AI stimuli, which requires no programming and relies on commercially available tools, increasing accessibility for researchers. Five studies (1,785 participants) directly compare visual stimuli generated using RAISE and iGenAI to stimuli generated by researchers and show that participants cannot differentiate them. Moreover, AI-generated stimuli satisfy the same manipulation checks in and replicate the effects of existing research. Three additional studies (N = 368) lend additional robustness, indicating that iGenAI and RAISE are valuable tools to complement traditional methods for producing visual experimental stimuli in advertising research.
科研通智能强力驱动
Strongly Powered by AbleSci AI