By Rimma Teper

The benefits of knowing thyself have been discussed by philosophers and scholars for centuries. As William Shakespeare wrote, “Of all knowledge, the wise and good seek most to know themselves.” Centuries earlier, Lao Tsu professed, “Knowing others is intelligence; knowing yourself is true wisdom.”

And most would agree. Not only are we motivated to understand our experiences and emotions1, we also happen to think that we are pretty accurate in our own self-views2. But how well do we really know ourselves? The symposium I chaired at APS last month focused on a very specific aspect of self-knowledge, namely affective and behavioral forecasting. This area of research tests how effective people are at predicting their emotions and behaviors. Perhaps not surprisingly, the general consensus among psychologists is that we are not very good at this. But why is this the case? And is there anything we can do to increase the accuracy of peoples’ forecasts? These are some of the questions that we explored in this symposium. We took a bottom-up approach and began by discussing the mechanisms that might result in faulty affective forecasts. We then explored the downstream consequences of these biases for judgments and behavioral predictions.

We know from past work that people overestimate future emotional states because of a bias called focalism. What this means is that, when forecasting their emotional reactions, people often focus their attention on specific aspects of an event or scenario, while ignoring others 3. So for instance, when imagining winning the lottery, people often fail to consider the many mundane features of life that go unchanged despite the heaps of money, causing them to overestimate happiness. When considering how fantastic it would be to live in California, people pay too much attention to the sunny weather and not enough attention to important factors like community safety and job opportunities4.

The data presented by Leaf Van Boven shed some new light on the relationship between attention and emotion. While it’s been long known that emotional stimuli are more likely to capture our attention than neutral stimuli, the effects of attention on perceived emotion had yet to be explored. In other words, can simply paying attention to something make it more emotional? Leaf and his colleagues tested this idea in a very simple, yet elegant design. They presented participants with images on a screen that varied in the number of time they rotated. Participants were instructed to pay attention to the images when they rotated in order to then indicate whether or not the image had changed after the rotation. In this way, the experimenters were able to manipulate the degree of attention that was allocated to each image. It turns out that the more frequently rotating images were rated as more emotional in nature, suggesting that simply paying attention to something causes it to adopt an emotional flavor. This might explain why focusing on the weather causes people to overestimate just how awesome it would be to live in California.

So what sorts of consequences might these attentional and emotional miscalibrations have? Eugene Caruso’s research suggests that the emotional weight we attribute to the various features of a situation can have a profound effect on our decisions and judgments. One phenomenon that illustrates this nicely is what Eugene refers to as the Temporal Doppler Effect.  In his studies, he finds that people perceive the future to be closer, and more salient than the past6.  Well, why should this matter? As it turns out, this temporal bias has all sorts of downstream consequences for the way that we reason about certain events. People judge moral transgressions, such as tax fraud, more harshly if they believe that the transgression is going to happen in the future than if they believe the transgression has already been committed7. In one study, participants believed that a car accident victim should be compensated a whole one million dollars more if they were told the accident had just occurred and that the victim’s suffering was yet to come. This was in comparison to participants who thought the victim had already experienced the suffering in the past. So it seems then, that the emotional weight we attach to the temporal characteristics of an event can sway our decisions in important ways (and perhaps unjustifiably so).

Having begun with a discussion of self-knowledge, readers are probably now wondering what all of this has to do with knowing thyself. It turns out that focusing on inconsequential aspects of future events results in inaccurate predictions of one’s own emotional reactions to those events. Furthermore, the inability to access the emotional flavor of a future situation may result in inaccurate predictions of future behaviors.

In Eva Buechel’s studies, forecasters are asked to imagine winning a prize that they have either a 10% chance or 90% chance of winning. Buechel and her colleagues find that people think they will be significantly happier after winning the prize they have a 10% chance of winning. The happiness of those who actually win the prize, however, doesn’t change as a function of the probability. Experiencers, in other words, are ostensibly focused on the prize itself, and thus don’t vary in their degree of happiness. Forecasters, on the other hand, seem to direct their attention to the probability of winning. Interestingly, when participants’ attention is diverted from the probability specification and onto an 8-digit string of numbers they are instructed to remember – the affective forecasting bias is attenuated8.  In other words, once participants are no longer focused on the probability, it seems to loose its emotional weight.

So far, the research presented here suggests that we systematically overestimate our future emotions. Is this always the case, however? The data that I presented suggests that the opposite is sometimes true – that there are times when people might underestimate just how emotionally intense a situation might be. This inability to access future emotional states then translates into behavioral mispredictions. In our research, we consistently find that forecasters overestimate how willing they would be to cheat on a math test for monetary gain. Those completing the math test cheat on 1 out of 15 questions on average, while forecasters predict cheating on almost one third of the questions. Importantly, this discrepancy can be explained by differences in physiological arousal between groups, such that participants actually completing the test are significantly more aroused than are forecasters. In other words, forecasters overestimate their likelihood to cheat because they underestimate how affectively intense the experience of cheating will be9. It seems, however, that there are ways of attenuating the discrepancy between forecasts and behaviors. For instance, when we make participants believe that they’re aroused (by making them think their heart is beating quickly), they produce moral forecasts that more closely match actual behaviors. We also find that greater trait emotional awareness is related to more accurate moral forecasts.

But how problematic are these forecasting errors, really? Why should it matter if we cannot accurately predict our emotional reactions, or behaviors (especially in cases where we seem to underestimate our goodness)? What I’d like to suggest is that “knowing thyself” in this way is in fact consequential. Being able to accurately forecast your emotional reactions to events is crucial for making good decisions. Knowing how anticlimactic buying that new TV would have been might have prompted you to spend your money elsewhere10. Finally, being able to properly forecast your behavior in future situations is crucial for effective planning and goal-setting. People are more likely to achieve goals when they make behavioral plans that link anticipated situations with specified behaviors11. One can imagine that this might be difficult to do if we have poor insight into the ways we will act. Lucky for us, there seem to be ways of attenuating these biases. And perhaps simply being aware of them can also do us some good.

  1. http://kslab.kaist.ac.kr/kse612/Festinger1954.pdf
  2. http://www.ncbi.nlm.nih.gov/pubmed/11642351
  3. http://psycnet.apa.org/journals/psp/78/5/821/
  4. http://www.psiexp.ss.uci.edu/research/teaching/Schkade_Kahneman_1998.pdf
  5. http://psycnet.apa.org/journals/psp/106/1/20
  6.  
  7.  
  8. http://psycnet.apa.org/journals/psp/106/1/20/
  9.  

  10.