An association is a link or connection.

A systematic error that may affect the results of a study because of weaknesses in its design, analysis or reporting.

In treatment comparisons, to blind someone means to keep them from knowing who received which treatment.

The basis of a claim is what supports it.

A confidence interval is a range that reflects the extent to which the play of chance may be responsible for a result from a study, such as an effect estimate. There is a high probability (usually 95%) that the actual effect estimate is within that range.

A claim is something someone says that could be right, but could be wrong.

A comparison of treatments is a study where researchers look carefully at the differences in what happens between groups that use different treatments.

An effect of a treatment is something that it makes happen.

A treatment is effective if it causes something to happen that is wanted.

Evidence is facts used to support what you believe or decide.

In a comparison of treatments, fair means not giving an advantage to one treatment over another.

A fair comparison of treatments is one where the only important difference between the groups that are compared is the treatments they receive.

A series of actions taken to treat medical conditions. This could include providing medications, conducting a procedure, or giving preventative treatment but also covers wider actions such as enacting environmental or behavioural changes.

In treatment comparisons, an outcome (or ‘outcome measure’) is something good or bad that can happen after a treatment and is measured in studies.

Diagnosis of a “disease” (e.g. through screening) which will never cause problems during a patient's lifetime

Unnecessary treatments (e.g. for a condition that causes no symptoms, and will go away on its own), or intensive treatments for a health condition that could be remedied with less intensive treatment.

The p-value is the probability (ranging from 0 to 1) that the results observed in a study could have occurred by the play of chance, if the treatment actually had no impact on the outcome.

Scientific work which has been evaulated by one or more independent professionals who specialise in the subject area.

Just believing that a treatment works can change how someone feels. This is called a placebo effect. Placebo effects are presumed to act psychologically through suggestion.

A placebo is a dummy or sham treatment that does not contain active ingredients, which has been designed to be indistinguishable from the active treatment(s) being assessed. It is used to blind participants and others involved in a study of treatment effects, and to reduce the risk of placebo effects.

When investigations are repeated with new participant groups and the results remain similar, these results are said to be reliable.

A side effect is a harmful or unpleasant effect of a treatment that is not planned.

A “statistically significant” result is unlikely to have happened by chance. The usual threshold for this judgement is a probability of less than 5% (0.05).

A treatment comparison is a study where researchers look carefully at the differences in what happens between groups that use different treatments.

A treatment effect is something a treatment makes happen.

A series of actions taken to treat medical conditions. This could include providing medications, conducting a procedure, or giving preventative treatment.

An unfair comparison of treatments is one where there are important differences between the groups that are compared besides the treatments they receive.

In a comparison of treatments, unfair means giving an advantage to one treatment over another.