Using large language models to generate silicon samples in consumer and marketing research: Challenges, opportunities, and guidelines

Abstract

Should consumer researchers employ silicon samples and artificially generated data based on large language models, such as GPT, to mimic human respondents' behavior? In this paper, we review recent research that has compared result patterns from silicon and human samples, finding that results vary considerably across different domains. Based on these results, we present specific recommendations for silicon sample use in consumer and marketing research. We argue that silicon samples hold particular promise in upstream parts of the research process such as qualitative pretesting and pilot studies, where researchers collect external information to safeguard follow‐up design choices. We also provide a critical assessment and recommendations for using silicon samples in main studies. Finally, we discuss ethical issues of silicon sample use and present future research avenues.

Large language models complement human participants through role simulation

“…Such misalignments are not merely theoretical or philosophical, but manifest in empirical response patterns. For example, LLM responses have been found to mischaracterize marginalized groups, as evidenced by out-group imitation rather than in-group description 13 , and misrepresent sampled groups, as demonstrated by upward bias in mean ratings of the Big Five personality traits 14 , as well as in other surveys and tests [15][16][17] . Beyond these shifts in average responses, LLMs also fail to capture the nuances and heterogeneity of human responses, producing flattened, oversimplified portrayals of various groups 13 .…”