This story is over 5 years old.

Watch this Scientist Give a Step-by-Step Takedown of Anecdotal Evidence

"The plural of anecdote is not data."

Does a seemingly positive personal experience with a health treatment qualify as evidence that it works? According to biological scientist Jonathan Jarry, the answer is no, and he explains why in this video.

"The plural of anecdote is not data," Jarry emphasizes. So even if a large body of people testify that a treatment such as reiki, or energy healing, worked for them, the collective anecdotes qualify only as "dirty data," he says.

Jarry points to examples of how patients suffering from cancer, the flu, or diffuse pain may successfully try reiki treatment—but the reduction or elimination of symptoms does not necessarily mean that the reiki was responsible.

A pool of anecdotal evidence, or "dirty data," is contaminated by variables that were not controlled for, he explains. In the case of the cancer patient, flu patient, and pain patient, the data could be inherently faulty based on factors such as whether the cancer patient was actually healed from simultaneous chemotherapy, whether the flu patient would have seen a natural decline in symptoms anyways, and whether the pain patient experienced a placebo effect due to the financial investment of reiki therapy.

Moreover, Jarry points out, "dead men tell no tales," meaning that anecdotal evidence that a health treatment works comes only from the survivors: A cancer patient for whom reiki failed wouldn't have lived to say so.

In contrast to dirty data, "clean data" is supported by rigorous scientific experimentation that minimizes variability and biases. A clean data set, says Jarry, is essential for clinicians to make the right decisions.