The journal "Nature Human Behavior" published a systematic review on the 20th, arguing that the most frequently recommended strategies for promoting happiness in the media - such as mindfulness and exercise - have a weak scientific basis.
The standard of high-quality scientific evidence in the field of psychology has changed over the past decade. Common research practices in the past, such as selective reporting or exclusion of certain participants, may increase the number of false positives in the results. Many psychologists now pre-register their studies, submitting specific methodological and analytical decisions in advance, and increasing the study sample size to improve the statistical power of the study.
A research team at the University of British Columbia in Canada investigated the empirical evidence behind some common happiness-enhancing strategies. They conducted an Internet-based media review and found the five most frequently recommended strategies for improving well-being: expressing gratitude, increasing sociability, exercising, mindfulness or meditation, and increasing exposure to natural environments.
The research team then searched the published scientific literature and found 494 papers detailing 532 studies involving the well-being benefits of one of the five strategies in a nonclinical sample. In nearly 95 percent of the trials based on being in nature, exercising, or participating in mindfulness meditation, the researchers found that the studies lacked sufficient statistical power to detect visible benefits. They identified only 57 studies that were pre-registered, or had well-powered experiments, that tested the effects of these strategies on subjective well-being in healthy individuals.
The team concluded that it is not clear whether these happiness strategies recommended by the mainstream media can really improve happiness, and more powerful pre-registration studies are needed to test these strategies.
(Originally titled "Scientific evidence that mindfulness improves well-being is insufficient")