Culture Changes the Meaning of the Message
Social-norm messaging does not reach a culturally neutral audience. The China–US food-waste experiment provides a focused test of that proposition by holding the basic normative information constant while examining whether cultural context changes what participants perceive and how those perceptions relate to intention. The useful judgment is not which country is “more social,” but where variation enters the pathway—from message to perceived norm, and from perceived norm to a stated plan to act.
Same messages, different pathways
The experiment recruited 1,449 university students: 563 from the United States and 886 from mainland China. Participants were randomly assigned to one of four combinations of descriptive and injunctive information. The descriptive manipulation varied the reported prevalence of food-waste prevention: one message said that about 80% of students had taken action, while the other said that only about 20% had. The injunctive manipulation varied social approval: one message said that most students considered preventing food waste very important and wasting food unacceptable; the other said that only a small number held those views.
This design matters because it does more than compare two populations after showing them vaguely different campaigns. It independently varies what students are said to do and what students are said to approve. The researchers also used the same cover story, “Waste Wise,” an on-campus sustainability program, and worked to make the English and Mandarin materials conceptually equivalent through translation and back-translation. The design therefore permits a more precise question: when the normative content is manipulated in the same logical way, does the cultural context change its effect?
The message checks worked as intended. Participants who saw the high-prevalence information rated food-waste prevention as more prevalent than those who saw the low-prevalence information. Participants who saw strong approval rated the behavior as more socially approved than those who saw weak approval. The two message manipulations did not produce an unwanted interaction in these perception checks, and the messages were judged similarly believable across conditions and countries. That matters because a cultural difference in outcome is easier to interpret when it is not simply a difference in whether one group found the message credible.
The main result appeared in the mediation pathway. Descriptive information did not directly produce a significant change in behavioral intention once the model accounted for perceived descriptive norms and other covariates. Instead, the message changed how prevalent students perceived food-waste prevention to be, and that perceived prevalence was associated with stronger intentions. The pathway was stronger in the Chinese sample at both stages. The same descriptive manipulation produced a larger change in perceived prevalence among Chinese participants than among US participants, and perceived prevalence had a stronger relationship with intention in China.
The injunctive model showed a parallel pattern. The approval manipulation changed perceived injunctive norms, and perceived approval predicted stronger intentions to prevent food waste. Again, cultural context moderated both the message-to-perception path and the perception-to-intention path. Chinese participants responded more strongly to the same approval manipulation in their perceived norms, and their perceived approval was more strongly associated with intention than it was for US participants. The direct effect of the injunctive message on intention was not significant once the perceived norm pathway was modeled.