Research literacy primer

A practice is evidence-based when there is enough high-quality research to support its use for a specific condition. It does not mean it works for everyone, and it does not mean it is the only thing that works. It means we have enough trustworthy data to say, with some confidence, that the practice is more helpful than placebo or usual care for certain people under certain conditions.

"Evidence-based" is also sometimes used loosely. A workshop brochure may say it is evidence-based because one small study found a modest benefit for a vaguely related outcome. That is not the same as a well-established treatment like EMDR for PTSD, which has multiple randomized trials, systematic reviews, and clinical guidelines behind it.

Not all research is equally strong. Think of evidence as a pyramid, with the strongest designs at the top and the weakest at the bottom.

• Systematic reviews and meta-analyses: researchers gather every study they can find on a topic and combine the results. These sit at the top because they summarize many studies rather than relying on just one.
• Randomized controlled trials (RCTs): participants are randomly assigned to receive the treatment or a control condition. This is the strongest design for testing whether the treatment itself caused the improvement.
• Controlled studies without randomization: participants are compared to a similar group, but assignment was not random. Still useful, but easier to bias.
• Case studies and testimonials: a single person or a small group reports what happened. These are stories, not proof. They can generate ideas, but they should not convince you a treatment works for everyone.
• Expert opinion: what a knowledgeable person thinks. It can be valuable, but it is still opinion, not data.

For cultural and spiritual practices, the pyramid looks different. Generational knowledge, community experience, and lineage-based practice are forms of evidence that do not fit neatly into clinical trials. They are worth respecting, but they answer different questions than a randomized trial does.

You do not need to read every word. Start with the abstract, which is a short summary. If it is behind a paywall, check PubMed Central for a free version, or look for a plain-language summary on the authors' website.

• Look at the question: What did the researchers actually study? Make sure the question matches the headline claim. A study on yoga for stress reduction is not a study on yoga for PTSD.
• Check the sample size: How many people took part? A study of 12 people might be interesting, but it is not a conclusion. Larger studies usually give more reliable results.
• Look for randomization and a control group: If participants were randomly assigned to a treatment group and a control group, the study is stronger. Without a control group, you cannot tell whether the practice helped, or whether people simply got better over time.
• Ask what was measured: Did they use a validated questionnaire, or did they ask people how they felt? Both matter, but the more objective and specific the measure, the stronger the evidence.
• Find the effect size: A study can say a treatment was statistically significant while the actual benefit was tiny. Look for numbers like effect size, percentage improvement, or how much symptoms dropped on a scale.
• Read the limitations: Good researchers list what their study cannot show. If the limitations section is missing or dismissive, be cautious.

Some claims are designed to sound scientific while actually being weak. Here are warning signs that a claim is being oversold.

• The phrase 'clinically proven' with no reference to a specific study, journal, or condition.
• A single small study is treated as the final word, especially if it was never replicated.
• The study was paid for by the people selling the product or training, with no independent replication.
• The claim jumps from one condition to another. A study on meditation for stress is not proof it treats depression.
• The words 'ancient' or 'natural' are used as if they mean 'scientifically proven.' Long history does not equal clinical evidence.
• The claim relies on testimonials, before-and-after stories, or influencer endorsements as primary evidence.
• The phrase 'studies show' is used without naming which studies, in what populations, and with what outcomes.

This is the most important sentence in research literacy. Just because two things happen together does not mean one caused the other. People who drink green tea may live longer, but that might be because people who drink green tea also exercise more, sleep better, and have more money for healthcare.

A randomized controlled trial is the best way to move from correlation toward causation. If people are randomly assigned to the practice, and they still improve more than the control group, it becomes much more reasonable to believe the practice itself caused the benefit.

Peer review means that other researchers in the same field read the study before it was published and raised questions or objections. It is a basic filter, not a guarantee. Peer-reviewed studies can still be flawed, biased, or unreplicated.

A study published in a peer-reviewed journal is more credible than a blog post or a sales page. But the real question is still the quality of the study itself: sample size, design, outcome measures, conflicts of interest, and whether the results have been replicated.

A conflict of interest does not automatically make a study wrong. But it does mean you should look more carefully. Common sources of bias in healing research include:

• The study was funded by the company selling the treatment, supplement, or certification.
• The author runs a training organization that certifies practitioners in the method being studied.
• The publication is a professional organization's own journal, which may be less likely to publish negative results.
• The study was too small or too short to detect harm, or it only reported positive outcomes while ignoring negative ones.

If you cannot find the funding source or conflicts of interest, that is also a reason to hold the claim lightly. Transparency matters.

Many practices, especially cultural and spiritual traditions, have not been studied in large clinical trials. That is a statement about the research economy and the priorities of science, not proof that the practice fails.

It is also not proof that the practice works. The honest position is: we do not know yet, or the evidence exists in a different form than a clinical trial. For serious conditions, use practices with stronger evidence as your primary support, and add others as complementary.

• Who funded the study, and what do they have to gain from the result?
• How many people were in it, and were they similar to me in age, condition, and life situation?
• Was there a control group? Was assignment random?
• What outcome was measured, and was the measure specific and meaningful?
• Has the finding been replicated by researchers without a financial stake?
• Was it published in a peer-reviewed journal, or is it only on a marketing page?
• Does the headline match the actual study question, or is it exaggerating?

• PubMed (pubmed.ncbi.nlm.nih.gov): free, searchable database of biomedical and clinical research. PubMed Central often includes free full-text articles.
• Cochrane Reviews (cochranelibrary.com): systematic reviews designed to help people make health decisions. Plain-language summaries are available.
• NCCIH (nccih.nih.gov): NIH center on complementary and integrative health with plain-language summaries of evidence for many practices.
• APA PsycInfo: useful for psychology and trauma-focused modalities, though full access may require a library login.
• Professional organizations for the modality: read them with the same critical eye you would bring to any advocate, but they often collect the best available research.

Placebo is real, and that is not a slur. A meaningful part of any healing practice, including licensed medicine, is the therapeutic relationship, hope, meaning-making, and the body's own capacity to settle. If a practice gives you a container for those, it may genuinely help even when the mechanism is unclear.

Evidence catches up slowly. Practices sometimes become well-supported decades after practitioners already knew they worked. EMDR is a classic example. Waiting for perfect evidence is a reasonable choice, but it is still a choice, not a neutral stance.