- Meta-analysis: This is a study comprised of the findings of other studies. The authors typically round up trials on a given topic and attempt to combine the findings through statistical analysis. This may expand the overall validity of findings by increasing the numbers and the diversity of the sample. It can be tricky to understand the methodology, though, and to be sure that like is matched with like.
Peer-reviewed: This means that an article was subjected to review by other scientists in the field before being accepted for publication. Peer-review is super important because it holds science to a high standard. An study that has been peer-reviewed is more likely to be free from major issues in its concept or design than a study that hasn’t been through the process. The peer-review process is often blinded, so neither author nor reviewer know who the other is. This practice helps to cut down on bias. Think of peer-review like a system of checks and balances: you can’t just say whatever you want, regardless of whether it’s valid, because other experts are looking over your shoulder. Peer-reviewed articles are likely to be published in academic, scholarly journals and less likely to be published in the popular press (magazines and newspapers) or on blogs.
- RCT (randomized controlled trial): Often called the gold standard of clinical evidence. The “controlled” part means that there are at least two groups: one that gets the treatment (intervention), and one that doesn’t (control). The control group could get a placebo, or a different, more standard treatment, or maybe nothing— but there needs to be something we’re comparing to. If neither the participants nor the person administering the treatment know whether they’re getting intervention or placebo, it’s called a double-blind RCT. This helps cut down on potential bias. The “randomized” part refers to how the researchers decided who was in the intervention group and who was in the treatment group, and its another way to cut down on bias. If participants are randomly assigned, then the two groups are likely to be equivilant. This is important because it makes it more likely that any difference they find is actually caused by the intervention. So, a well-done RCT is generally a good source of evidence about whether an intervention works under the conditions it was tested in. Note, though, that those conditions might not be anything like real-life conditions, and the participants in the study might not be anything like you. There are also a few cases where RCTs might not be much use. Some things are hard to test under highly controlled conditions— like nutrition. It can also be hard to see long-term effects in RCTs since they are expensive to conduct and might have short follow-up periods.
- Systematic Review: A review of all the current published scientific evidence on a topic. These are typically conducted as a round-up, state-of-the science kind of project. The authors usually search several databases in an attempt to collect all the evidence on a particular topic. They may limit their search to articles from a certain time period, a specific population, or a particular kind of intervention. They often then choose articles they deem to be high-quality (such as RCTs) and exclude inferior articles. The result is a summary of what the scientific community knows about a topic.
Suggestions of other terms that need definition or explanation? Comment below.