Verified by 
Key points
- The field of Positive Psychology has argued there are strategies you can use to become happier.
- But a new paper argues that there are limitations in the majority of previous experiments on happiness.
- This means that we may need to be more cautious about what can actually increase happiness.
One of the key findings from the field of positive psychology is that there are small strategies — like exercise, meditation, and keeping a gratitude journal — that can have a lasting impact on our happiness. Over the past thirty years, psychologists have conducted hundreds of experiments to try to find which of these strategies people should apply in their everyday lives to become happier.
The results from these studies have been shared with the public through self-help books, podcasts, and TED talks, making a large cultural impact. For example, there are over a hundred TED talks on the topic of happiness, and many of these talks have been viewed by millions (or tens of millions).
The science of how to become happier has influenced millions of people around the world, and created a multi-billion dollar happiness industry.
There’s only one problem. According to a recent article from Dunigan Folk and Elizabeth Dunn (2023), there are limitations in the quality of evidence in the vast majority of happiness research.
What Makes an Experiment Trustworthy?
Psychologists conduct and publish many experiments, but not all of these experiments are created equal. Some are more likely to replicate (in other words, produce the same results in a new study in the future) than others. Unfortunately, there is no simple strategy for telling whether an experiment will replicate or not.
But Folk and Dunn (2023) argue that there are two important features of experiments that give us a hint as to whether the results are likely to replicate:
Pre-registration
When a study is pre-registered, that means that the authors of the study started out with a plan for how they would conduct the study and then analyze and interpret the data. Pre-registration is the scientific equivalent of “calling your shot”. Arguably, when studies aren’t pre-registered, there is a risk that researchers will change their plans after the fact to present the results of studies in a more favorable light. When there’s no plan ahead of time, this means there’s more wiggle room in how analyses are conducted (and this wiggle room usually nudges researchers in favor of finding a significant result).
Of course, we should not be immediately dismissive of studies that aren’t pre-registered. But this does mean that many current psychologists will interpret results more favorably when they see that a study was pre-registered.
Critically, Folk and Dunn found that only 55 out of 494 studies (about 11%) were pre-registered. This is not surprising, as the technique of pre-registration has only become commonplace in the past decade.
Statistical Power
When psychologists talk about the concept of statistical power, they are interested in whether an experiment had a large enough number of participants to reliably detect an effect. When an experiment has higher power (and a larger number of participants) we can be more confident that the results represent a real effect (that we can observe again in the future) and are not just a lucky coincidence. The exact number of participants depends on the situation; in happiness research, you may need hundreds (or thousands) of participants in order to have what’s considered a “high-power” study.
This does not mean that larger studies are always better — if your methods are flawed, it doesn’t matter how many participants you recruit. But it does mean that all else being equal, having a larger sample size is a good thing.
Folk and Dunn argue that only 6 out of 494 articles (less than 2%) had sufficient statistical power. Like pre-registration, psychologists have begun increasingly concerned with statistical power in recent years.
Can You Trust the Science of Happiness?
Folk and Dunn (2023) found that there were very few experiments on happiness that were both pre-registered and sufficiently powered.
This doesn’t mean that we should throw out all of the existing research on how to become happier.
But it does mean that we should be more skeptical about what we thought we knew. And it means that there is a real risk that authors and speakers have over-sold the value of techniques to increase happiness.
As is often the case in behavioral science, more data is needed.
References
Folk, D., & Dunn, E. (2023). A systematic review of the strength of evidence for the most commonly recommended happiness strategies in mainstream media. Nature Human Behaviour, 1-11.
