Why do online reviewers publish negative ratings despite positive experiences?

Reviewing online is communication to a multiple audience — people who liked the product and people who didn't, simultaneously. Schlosser's experiments isolate a self-presentational mechanism that distinguishes posting from private rating. After reading a negative review, posters lower their public rating relative to the no-review and positive-review conditions, even when their personal experience with the product was favorable. Lurkers, who rate privately, show no such effect.

The mechanism: negative evaluators are perceived as more intelligent, competent, and expert than positive evaluators (Amabile 1983). Reading a negative review primes posters to worry their own positive opinion will look indiscriminate or low-standards, so they hedge downward to seem more discerning. The effect is asymmetric — positive reviews don't trigger the equivalent worry because positive reviewers don't carry the intelligence-signaling effect.

Posters also acknowledge multiple sides of the issue more than lurkers do, but they do not integrate the sides — they hold them as parallel parallel claims rather than synthesizing them. Lurkers, freer of social pressure, are more likely to integrate multiple viewpoints into a single coherent judgment.

This contradicts cognitive-tuning research (which predicts polarization toward attitudes), Grice's cooperative-principle maxims (which posters violate by suppressing their genuine positive experience), and the assumption that anticipated social interaction prevents negativity bias. The findings are specifically about multiple-audience public communication. The implication for recommender systems: aggregated ratings in multi-audience platforms systematically understate true average satisfaction for products that received any negative review, because every subsequent positive-experiencer adjusts their rating downward in public.

Source: Recommenders General