Levenson, Ekman, Heider and Friesen (1992)

Chapter 1 endnote 24, from How Emotions are Made: The Secret Life of the Brain by Lisa Feldman Barrett.
Some context is:

This alternative explanation is borne out by their later experiment with an Indonesian tribe, the Minangkabau of West Sumatra. These volunteers had less understanding of Western emotions and did not show the same physical changes as Western test subjects; they also reported feeling the expected emotion much less frequently than the Western subjects did.

Ekman, Levenson and their team attempted to replicate their 1983 findings, this time testing Western college students and community volunteers, as well as people from a non-Western culture, the Minangkabau of West Sumatra. Test subjects were given muscle-by-muscle instructions and coaching to pose their faces into the stereotypes of the basic emotion method. In addition, some participants viewed films or relived (remembered) prior emotional episodes.

From one point of view, these studies were extraordinary. When you think about how many things can go wrong in a laboratory, let alone when traipsing across the globe into other cultures, carting around complicated machinery in uncontrolled conditions, with test subjects who were completely unfamiliar with Western experimental methods… these experiments were a virtuosic achievement.

Nevertheless, the results fall drastically short of what the classical view predicts. The published paper claims to have found evidence for emotion fingerprints, but the studies really did not replicate the 1983 results all that well, and there was too much variability to confidently claim evidence for emotion fingerprints. The Western subjects reported experiencing the desired “target” emotion only about half of the time in the best of the studies, and less than a third of the time in some of the others, while the Minangkabau rates were even lower, violating a major tenet of the classical view of emotion.

There was an alternative explanation for the findings, as well. Test subjects were instructed on the facial poses they made. Given what we now know about simulation and how language comprehension works, these instructions very likely brought emotion concepts to mind. That is, the information allowed Western subjects with conceptual knowledge to simulate emotion. In fact, people can correctly identify most of the target emotions from such instructions.

One reason that the Minangkabau evidence is weaker, then, is that the Minangkabau might not have shared the same emotion concepts as Westerners, leaving them less susceptible to simulation. And so, compared to more Western subjects, they would not have the conceptual knowledge about which emotion they were “supposed” to be experiencing. In fact, evidence from experiments with other test subjects from remote locations, who live in relative isolation from U.S. cultural practices and norms, do not appear to share many of the same emotion concepts as Westerners.