The 9-Year-Old Who Beat the Experts
Opening
By the mid-1990s, Therapeutic Touch was being practiced in hospitals across North America. More than a hundred thousand nurses had been trained in the technique. Courses were offered at universities. The method had been approved as a nursing intervention by several state boards and had found its way into mainstream healthcare settings from New York to California.
The premise of Therapeutic Touch is that trained practitioners can detect and manipulate the human energy field, an invisible biofield that surrounds the body, using their hands. By sensing imbalances in this field and redirecting its flow, practitioners claimed to reduce pain, accelerate healing, and promote wellbeing. Patients reported feeling calmer. Hospitals were persuaded. Insurance companies reimbursed it.
There was, however, one small thing no one had done.
No one had tested whether practitioners could actually detect the human energy field.
That was about to change. The person who would finally test it was nine years old.
The Science Fair Project That Made It Into JAMA
Emily Rosa was in fourth grade in Loveland, Colorado, when she designed what would become the most unusual entry in the history of the Journal of the American Medical Association. Her mother, Linda Rosa, was a registered nurse who had been skeptical of Therapeutic Touch for years. James Randi, the professional skeptic and magician, had been publicly challenging practitioners to prove their claims under controlled conditions. Emily, looking for a science fair project, decided to give them the chance.
Her design was elegant and devastatingly simple. She built a screen out of cardboard, with two holes cut at the bottom. A Therapeutic Touch practitioner would sit on one side of the screen with their hands through the holes, palms up. Emily would sit on the other side. She would flip a coin to decide which of the practitioner's hands she would hold her hand above, then lower her hand silently into place. The practitioner's task was to identify which hand was in the proximity of Emily's energy field.
If Therapeutic Touch practitioners could truly detect human energy fields, they should be able to identify the correct hand at a rate better than chance. Fifty percent correct would be expected by guessing alone. If the energy field was real and detectable, scores should be significantly higher.
Emily ran the experiment first for her school science fair, then extended it with additional practitioners for the formal study. Twenty-one practitioners participated, performing a total of 280 trials.
The Result
The practitioners identified the correct hand 44 percent of the time.
Not better than chance. Slightly worse. The human energy field, as a detectable phenomenon, did not show up in the data. Under controlled conditions, with the practitioners unable to see Emily's hands or receive any visual or auditory cues, their claimed ability to sense the biofield vanished entirely.
The paper was submitted to JAMA. The editors subjected it to peer review. The reviewers, including a statistician, found the methodology sound. JAMA published it in April 1998. Emily Rosa was eleven years old at the time of publication, making her the youngest person ever to be the first author of a paper in a peer-reviewed medical journal. The record stands.
The paper noted that the study had limitations: the sample size, while adequate for the basic question, was not enormous; the practitioners were self-selected volunteers rather than a random sample; and additional variables like the direction of Emily's palm or the distance above the hands were not systematically controlled. But on the central question, the finding was clear. Under the simplest possible blind test, practitioners could not do what they claimed to do.
The Question the Result Raises
Emily Rosa's study did not end the practice of Therapeutic Touch. Practitioners responded that the conditions were not ideal, that a child's energy field was not representative, that the screen interfered with the biofield, that their ability only manifested in clinical contexts. The practice continued in hospitals for years afterward.
But the study raised a question that is more important than the specific finding about Therapeutic Touch: how does something like this become established medical practice without anyone first checking whether it works?
The answer illuminates something important about the difference between authority and evidence. Therapeutic Touch spread through professional channels, through training programs, nursing journals, and hospital endorsements. Each of these institutions lent it credibility without independently verifying the foundational claim. The energy field was assumed to exist because practitioners reported sensing it and patients reported feeling better. Nobody thought to ask: can practitioners actually detect the field, separate from everything else happening in a clinical encounter?
A nine-year-old with a cardboard screen thought to ask. It took under an afternoon to get the answer. The experiment required no laboratory, no funding, no institutional affiliation. It required only the insight that the basic claim could be tested and the willingness to test it.
This is the real lesson of Emily Rosa's study: not that Therapeutic Touch failed a test, but that such an obvious test had never been run. Expert consensus, professional training, and widespread adoption are not substitutes for a basic empirical check.
Expert Illusion / Untested Assumption: Practices and beliefs can become widely accepted through professional adoption, institutional endorsement, and reported subjective experience without anyone testing the underlying empirical claim. The perception of expertise, authority, and widespread use is not evidence that a mechanism works as described. The willingness to apply the simplest possible empirical test is often the only thing standing between assumption and knowledge.
What this means for a regular Tuesday
Ask what would have to be true for this to work, and whether that has been tested.
Before accepting any claimed mechanism, whether in business, medicine, management, or technology, ask what specific testable prediction the mechanism makes. Therapeutic Touch predicted that practitioners could detect energy fields; nobody tested the detection. Many management practices, productivity systems, and medical interventions make specific empirical claims that have never been directly checked. The question is not whether the outcomes are reported as positive, but whether the mechanism is actually operating.
Professional consensus is evidence, but not proof.
The fact that a practice is taught in professional training, endorsed by institutions, and widely adopted is genuine information. It is not nothing. But it is not a substitute for asking whether the practice has been empirically tested under conditions that could actually produce a negative result. Practices can survive and spread for reasons entirely unrelated to their efficacy: they feel intuitive, they produce strong placebo responses, they are simple to teach, they fill a gap in a service offering.
Simple tests often work.
Emily Rosa's experiment cost nothing and took a few hours. It required only the observation that the core claim was testable and the design of a condition under which a negative result was possible. Many assumptions in organizations and personal life are similarly testable and similarly untested. The barrier is not technical difficulty; it is usually the assumption that something widely practiced must already have been verified.
How AI can help here
Use the pushback-oriented setup from The Man Who Robbed Banks With Lemon Juice in the main book for prompts that challenge your first instinct.
Use AI to pressure-test any practice, belief, or recommendation you are relying on. Ask it to identify the specific empirical claim being made and whether it has been tested under conditions that would have permitted a negative result.
I rely on or am considering adopting the following practice or belief: [describe]. The claim behind it is approximately: [describe what it is supposed to do or how it is supposed to work]. Help me test this claim rigorously: (1) What specific, observable prediction does this mechanism make? (2) Has that prediction been tested under conditions where a negative result was possible? What does the evidence show? (3) Is there a simple test I could run myself, or a piece of evidence I could look for, that would distinguish "this works as claimed" from "this seems to work for other reasons"?Before adopting any widely-used professional practice, use AI to investigate whether the widespread adoption is based on tested efficacy or on diffusion through professional channels without direct empirical validation.
The following practice is widely used in [field / industry / organization]: [describe]. It is described as working because: [describe the claimed mechanism]. I want to know whether the underlying mechanism has been directly tested, or whether adoption was primarily driven by professional diffusion, intuitive appeal, or reported subjective outcomes rather than controlled evidence. What does the research actually show? Is the specific mechanism validated, or just the general outcome? Are there any studies that specifically tested whether the claimed mechanism is operating rather than something else?When reviewing expert advice, use AI to help you design the simplest possible test of the key claim, one that a negative result would actually be visible in.
An expert has advised me that [describe the advice or claim]. I want to verify this independently before acting on it. Help me design the simplest possible test of this claim: What would I observe if the claim were true? What would I observe if it were false? What is the minimum I would need to do to distinguish between these two cases? Are there existing data sources that would already answer this question?
References
Linda Rosa, Emily Rosa, Larry Sarner, and Stephen Barrett. "A Close Look at Therapeutic Touch." Journal of the American Medical Association, Vol. 279, No. 13, 1998, pp. 1005-1010.
The original paper. Emily is listed as second author at age 11.
James Randi. "Therapeutic Touch: What Could Be the Harm?" Skeptical Inquirer, Vol. 20, No. 4, 1996.
Randi had been offering a substantial prize to anyone who could demonstrate paranormal abilities under controlled conditions. No TT practitioner ever successfully claimed it.
Dónal O'Mathúna and Walt Larimore. Alternative Medicine: The Christian Handbook. Zondervan, 2001. Chapter on Therapeutic Touch summarizes the evidence base, including the Rosa study, and is representative of subsequent clinical assessments.
The asterisk
The Therapeutic Touch community's response to the Rosa study was largely to question the methodology rather than update the practice. Common objections included: that the screen interrupted the biofield; that practitioners perform better in genuine clinical contexts; that Emily's hand was not a normal adult hand; that the coin-flip protocol introduced artifacts. Several of these are legitimate points about boundary conditions, but none of them change the fundamental finding: under the simplest possible blind test, the claimed detection ability was not present.
A broader research literature on energy healing and biofield therapies exists, and the evidence base is generally weak. Systematic reviews tend to find either no effect or effects indistinguishable from placebo or relaxation responses. This does not mean patients do not feel better after Therapeutic Touch sessions: they often do. But "patients feel better" is consistent with multiple explanations, including touch, attention, relaxation, and expectation effects, that do not require a detectable biofield.
The Rosa study's deeper significance lies in its demonstration of a general failure mode: practices with no mechanism check, no falsifiable prediction, and no controlled test can survive and spread for decades in professional settings that nominally value evidence. The lesson applies well beyond Therapeutic Touch.
The experiment did not require a laboratory, a grant, or a decade of training. It required the question that nobody with a decade of training had thought to ask.