Stories, Statistics, and Memory*.

This article investigates how different types of information—stories (qualitative accounts of individual instances) versus statistics (quantitative summaries)—affect selective memory and the persistence of belief updates over time. Through a series of controlled experiments, it finds a pronounced "story-statistic gap": while statistics initially influence beliefs more strongly, their impact decays by about 73% after one day, whereas the effect of stories decays by only 32%. This difference is explained by a model of cue-dependent memory, where stories contain qualitative content semantically related to memory cues, enhancing recall, while statistics lack such associations and are more prone to forgetting and interference from similar memories. The study further shows that adding uninformative qualitative content to statistics improves their recall and reduces belief decay, and that memory failures—rather than partial forgetting—primarily drive the story-statistic gap. These findings have implications for communication strategies in media and policy, suggesting that qualitative narratives help information persist in memory and influence beliefs more durably than statistics alone.