General information for adults. Not medical or psychotherapeutic advice; not a substitute for examination, diagnosis or treatment.
What question is being studied?
Research may examine benefits, unwanted effects or experiences. The appropriate design depends on the question. A survey and a treatment comparison provide different kinds of answers. [1]
Why does a fair comparison matter?
Randomised controlled trials allocate participants to groups by chance, then compare outcomes. This aims to reduce systematic differences between groups. [1]
Is the label “study” enough?
Conduct matters too: participant selection, dropouts, measurements and possible biases can affect reliability. It also matters to whom results apply. [2]
What is a systematic review?
It finds and assesses research on a defined question using explicit methods. Cochrane explains how studies are brought together. A review’s usefulness depends on its methods and the research available. [3]
An invented numerical example
Suppose 40 out of 100 people in one group and 30 out of 100 in another report improvement. The difference is 10 percentage points. Those figures alone cannot establish whether treatment caused it or how important it is to people. This is an arithmetic example, not a depression study result.
What does this mean for one person?
Research evidence, professional expertise and individual circumstances need to be considered together. An average cannot reliably predict one person’s outcome. [4]
READ FURTHER & CHECK THE SOURCES
Sources and context
How should I interpret these sources?
Source notes identify the provider, their use in this article and known limitations. The source-comparison date below is not the original publication date. No update date is assumed where none is available. General findings cannot automatically be applied to an individual; independent medical review of this draft is pending.
- IQWiG: Types of studies (German-language source) ↗
Provider / domain: www.gesundheitsinformation.de
Methods explainer on research questions, comparison groups and random allocation. Accessed: 29 October 2026.
- IQWiG: How studies are assessed (German-language source) ↗
Provider / domain: www.gesundheitsinformation.de
Methods explainer on bias, fair comparisons and applicability. Accessed: 29 October 2026.
- Cochrane: About us ↗
Provider / domain: www.cochrane.org
The organisation explains its systematic reviews; not evidence for a particular depression treatment. Accessed: 29 October 2026.
- IQWiG: Principles of evidence-based medicine (German-language source) ↗
Provider / domain: www.gesundheitsinformation.de
Methods explainer on combining research, professional expertise and personal circumstances. Accessed: 29 October 2026.
Research and source comparison: 29 October 2026. Prepared with AI assistance. Editorial responsibility: Mario V.W.B.R. Obst. Independent medical expert review is pending. Independent medical expert review has not taken place.