The predictive utility of word familiarity for online engagements and funding
https://www.pnas.org/content/118/18/e2026045118
How well do simple or complex language patterns predict meaningful behaviors in the field?
We used nearly 1.1 million datapoints across 12 samples to demonstrate that language complexity is a positive or negative heuristic depending on instrumental goal activation (e.g., if effort in a task is associated with value).
We substantiate the simpler-is-better hypothesis that suggests without instrumental goal activation, common words associate with favorable outcomes in social media, academia, and entertainment settings.
With instrumental goal activation, however, complexity leads to more favorable outcomes in the form of money for charitable giving campaigns and NIH grants.
Our first study examined whether the simpler-is-better hypothesis would obtain support in natural settings. We examined news attention (e.g., Twitter), social media (e.g., Reddit), science (e.g., PLoS One), and public entertainment speeches (e.g., TED talks). Together, these results consistently revealed, as people search these platforms, texts that use common words capture attention more than complex texts.
Our second study identified settings where a difficult experience would be valued and, thus, rewarded. According to Labroo and Kim (21), “…when trying to reach a goal, people must ask themselves, ‘Is this object any good for accomplishing my goal?’ and in this situation, an ‘instrumentality heuristic,’ or the naive belief that effort signals instrumentality, becomes pertinent.”
Thus, when engaging in an instrumental task such as donating to a cause or awarding a grant, people want to feel as though they exerted effort because, naively and automatically, this effort feels diagnostic within this context (22, 23). We specifically examined whether language complexity within charitable giving and grant funding would be positively associated with higher monetary rewards. Consistent with this premise, and again across multiple settings, we found that texts with more uncommon words were funded more than texts with more common words.