A kind of quasi-experiment where the investigators do not assign participants to the different conditions.
"A natural experiment takes advantage of differences in conditions that already occur, rather than being based on experimental manipulation… This may be especially useful where researchers are interested in the possible detrimental effect of some condition, and it would be unethical to create that condition and assign participants to it to test the effect (consider, for example, a study to find out if victims of bullying make less progress in their science classes – such a study would look to – sensitively – enrol existing students identified as victims rather than experimentally create new victims). Sometimes 'natural experiments' are possible due to some particular set of circumstances that happen to provide the type of comparison researchers are interested in studying." (Taber, 2019, p.85)
Read about 'Natural experiments'
