The science of happiness sounds great. But is the research solid?

Listen to Short Wave on Spotify, Apple Podcasts and Google Podcasts.

How do we really get happier?

In a new review in the journal Nature Human Behavior, researchers Elizabeth Dunn and Dunigan Folk found that many common strategies for increasing our happiness may not be supported by strong evidence. In fact, almost 95% of experiments on three common strategies—spending time in nature, exercise and engaging in mindfulness/meditation—did not hold up to even the most basic of current best practices for showing psychological effects.

The Replication Crisis

In their review, Dunn and Folk said recent changes in the way scientists conduct research have led their team to re-examine previous work in the field of psychology.

Around 2011, scientists faced a big challenge: the Replication Crisis. Researchers found that the results of many studies couldn't be replicated because they had too few subjects, used flawed methodologies or over-analyzed data so it showed statistically significant results.

Suddenly, multiple popular studies that seemed like they were backed by strong evidence were thrown into question.

Since then, many researchers have altered how they conduct experiments to ensure that their evidence really does back up their conclusions. But Dunn and Folk also wanted to look backwards at previous happiness research to see how research in their field held up against updated standards.

Setting the baselines

Before they could begin reviewing thousands of papers, Dunn and Folk had to identify their criteria for solid research.

First, they considered statistical power, which is the probability that researchers will detect an effect when one exists. Higher statistical power allows researchers to be more confident in the results of a study. This also has to do with the study's sample size, or how many people participated. Researchers today have realized that studies need to be a lot bigger than they historically were to establish that a psychological effect like increased happiness exists.

"If we have studies that have samples that are too small, we can end up failing to detect effects that really exist," Dunn said. "But we can also end up with a really high proportion of false positives where we think we found something, but it's not real."

Next, they looked at whether a paper was pre-registered. This means that scientists state exactly what they plan to do and how they're going to test their hypothesis, then make those intentions publicly available before they start the experiment and analyze the data.

"In the absence of pre-registration, researchers can kind of shoot darts at the wall and then draw the bull's eye on afterward. What pre-registration does is to prevent us from doing that," Dunn said. "We have to draw bull's-eye, shoot our dart and then be pretty honest and transparent about whether it hit."

Finally, the team had to decide on one definition of happiness to use in assessing papers for their literature review. They used a commonly accepted academic definition of happiness: subjective well-being (SWB). This measure includes two components: an emotional one, like feeling more positive emotions than negative emotions; and a cognitive one, like being satisfied with one's overall quality of life.

Happiness literature, reviewed

Dunn and Folk conducted a media search to identify the five strategies most commonly recommended for boosting mood: gratitude, exercise and physical activity, spending time in nature, meditation and mindfulness and social interaction. They screened thousands of studies that investigated one of these strategies' effects on SWB, narrowed down their results and ended up with 57 papers that met their criteria.

Here's what they found:

• Gratitude In their review, Dunn and Folk found some evidence supporting that writing messages of gratitude provides short-term increases in positive attitude.

• Social Interaction They also found evidence supporting that talking to strangers boosts mood, and that acting more extroverted can boost happiness. However, Dunn said the body of research on the topic is relatively small. 

• Mindfulness and Meditation For the mindfulness and meditation studies, the team found one pre-registered study on the topic which failed to support any benefits. Dunn said the team found eight experiments with sufficient statistical power. Of these eight, six showed that mindfulness-based stress reduction programs could increase happiness, but Dunn said they may have confounded social interaction with mindfulness practice. 

• Exercise The literature on exercise and physical activity also had relatively few studies that met the criteria for rigorous research. They found no pre-registered studies on the topic, but 12 well-powered ones. These studies found that people felt happier after a single bout of exercise, but did not provide evidence for long-term increases in happiness. 

• Time in Nature Finally, they examined the impact of nature exposure on SWB. After considering sample size, pre-registration and statistical power, the team found relatively little evidence supporting that time in nature contributes to happiness. 

Bonus review! For a glass half-full

Dunn and Folk conducted a follow-up literature review that will be published in the Annual Review of Psychology. They found evidence that multi-component interventions, or using more than one of these strategies at the same time, may increase happiness. The authors also saw evidence that receiving money can boost mood in people with fewer resources—and that giving money away can increase happiness for the giver.

Overall, Dunn said this lack of strong evidence in the field does not necessarily suggest these strategies are ineffective. Rather, they indicate a need for further research and for more caution by scientists in providing public guidance.

"I want to be really clear that, you know, if you're somebody who starts every day with a three mile run and it leaves you feeling happy and positive about life, by all means, continue it," she said.

Want to hear Dunn read the paper? Check it out here.

Listen to Short Wave on Spotify, Apple Podcasts and Google Podcasts.

What science story do you want to hear next on Short Wave? Email us at shortwave@npr.org.

This episode was produced and fact-checked by Rachel Carlson and Brit Hanson. It was edited by Rebecca Ramirez. The audio engineer was Josh Newell.