Understanding Evidence: Trials, Placebo Effects and Types of Research

Why types of research matter

Not every study answers the same question. NCCIH's "Know the Science" guide describes basic research, translational research, clinical trials, and systematic reviews and meta-analyses. [1] Knowing which kind you are reading tells you how much weight it can carry.

The main kinds of research

• Basic research asks questions about theories or concepts and tests hypotheses to improve scientific knowledge. It is the first step in any research. [1] Findings are labeled preclinical when they come from cells or animals.
• Clinical trials are done with people. NCCIH says they give the clearest information about whether a treatment or lifestyle change is effective and safe in humans. [1]
• Systematic reviews and meta-analyses examine many studies on the same topic. When many studies reach the same conclusion, the results are more likely to be reliable and valid. [1]

What a controlled trial is

In a controlled clinical trial, researchers compare the effects of different treatments in groups of participants who are as identical as possible in all other respects. [2] Participants are placed in an experimental group or a control group, which gets standard care or an inactive substance. [2] The commonly cited gold standard is the randomized, placebo-controlled trial. [3]

Blinding and bias

If either patients or investigators know which treatment a patient received, it may affect their impression of whether the patient improved. [4] In a double-blind study, neither the participants nor the researchers know who is in which group. [2] NCCIH advises checking whether a trial was blinded or masked. [4]

Many therapies are hard to blind. It is difficult to hide from a participant whether they did yoga or sat in a waiting room, and a therapist usually knows what they are doing. This is a real limit on how much those trials can show, and the studies often use a waitlist or usual-care comparison instead. Some trials use a "sham" (simulated) treatment as the comparison, and others compare with usual care or no treatment, so check which comparison was used before comparing results across studies. [5]

The placebo effect

NCCIH defines the placebo effect as a beneficial health outcome resulting from a person's anticipation that an intervention will help. [3] How a provider interacts with a patient may also bring about a positive response independent of any specific treatment. [3] NCCIH-supported research has suggested that placebos may be helpful even if patients know they are receiving them. [3] This means a person can feel better with a therapy that has no specific effect of its own, and a good trial is built to tell the two apart. A placebo response is real as an experience; it does not by itself show the treatment works.

Study size

The number of participants affects what the results mean. Small samples may produce inconclusive results and are more likely to produce results due to chance. [6] One small positive trial is therefore a reason for further research, not a final answer.

Putting it together: a reading checklist

Related pages

Apply these ideas with how to evaluate a therapy claim. For how much research exists on each approach, see therapies research and evidence. Evidence for an individual therapy is on its own page, for example yoga evidence by outcome.