Emotional states can be fleeting and somewhat inexplicable — you can feel great one minute and down in the dumps the next, sometimes for no apparent reason. It follows, then, that opinions based on emotion are likely to be equally fleeting: if you’re in a bad mood when you take part in a survey or review a product, then surely the attitudes measured and recorded will be just as transient too.
But according to a series of studies by Matthew D. Rocklage from the University of Massachusetts Boston and Andrew Luttrell from Ball State University, this isn’t actually the case. Instead, they report in Psychological Science, attitudes based in emotion are actually more stable: the more emotional an opinion, the less it changes over time.
In the first study, participants were asked to think of three gifts they had recently received, before selecting their attitude towards the gift from a list of adjectives, some overtly negative, some overtly positive and some neutral, such as “amazing”, “boring”, “terrifying”, or “valuable”. Participants listed gifts ranging from electric toothbrushes to Star Wars figurines.
One month later, participants were asked to think of the same gifts and again select adjectives that represented their feelings about them. After the second part of the study, the adjectives chosen by participants were coded for positive or negative valence, extremity, and emotionality. (Although these may seem similar, emotionality relates to how much an attitude is based on emotion, while extremity measures the extent to which an attitude is positive or negative; “outstanding”, for example, has high emotionality but low extremity.)
Those participants who chose more extreme adjectives, whether positive or negative, were less likely to see a change in the valence of the adjectives used to describe their gifts at the second time point. Similarly, the more an attitude was based on emotion, the less it changed too. A second study, which looked at attitudes towards brands, also found that emotionally-based attitudes changed less over time.
The third study looked at attitudes in a more naturalistic setting: reviews of products posted online. The team obtained all reviews for all restaurants in Chicago over a period of twelve years, looking only at reviewers who had posted more than one review of the same establishment. The team then analysed both the emotional valence of the reviews and measured any differences in the number of stars the reviewers gave the restaurants at each time point.
As in previous studies, positive emotionality consistently predicted less change in attitude across time, though negative emotionality did not. Positive extremity also predicted less change in attitude, while negative extremity predicted more.
A final study looked at whether exposure to messages designed to evoke emotions actually increase the likelihood of people developing fixed attitudes. To do this, the team assigned participants to two conditions. In one, they saw a message about a fictional aquatic animal called the “lemphur” designed to elicit high emotion, reading about a touching underwater encounter between the creature and a diver. In the low emotion condition, participants read a fact-based message about the lemphur similar to an encyclopedia entry.
After reading the text, participants indicated their attitude towards the animal, selecting from the same list of adjectives used in the first study. In follow-up studies over the next few days, participants selected adjectives again.
Those in the high-emotion condition were, unsurprisingly, more likely to indicate a more emotional response to the animal than those in the low-emotion condition, and also had a more extreme response. Those in the high-emotion condition also saw less change in their attitude towards the creature over time.
Overall, emotional responses were related to more fixed attitudes. Notably, positive emotionality had a particularly strong effect, which may be a useful to know for the creation of public health messaging or other attempts at attitude change — inducing positive emotions, rather than negative emotions like shame, may be more beneficial. Whether positive emotions have a similar effect on actual behaviour, rather than just attitudes, remains to be seen.