Interventional (experimental) studies see if an intervention that purposefully changes some aspect of a subject’s status (e.g. ingesting a drug) affects an outcome (e.g. recovery from disease). To ensure interventional studies are valid, randomization and blinding is often utilised, with randomized control trials (RCTs) being the gold-standard of clinical trials:

• Random allocation: Subjects are randomly allocated to different treatment groups so that baseline characteristics of participants in each group is as similar as possible. This eliminates selection bias (e.g. sicker individuals may consciously or unconsciously be allocated to what the allocator believes is the better treatment).
• Single blinding: participants do not know what intervention they get. This eliminates the placebo effect where a study participant’s belief in a treatment may, by itself, produce positive effects (vice versa is the nocebo effect)
• Double blinding: Single blinding plus investigator also does not know who gets what intervention. Further eliminates observer bias where a researcher’s own bias may subconsciously affect study participants.
• Triple blinding: Double blinding plus statistician (who does data handling) does not know allocated intervention groups