I'm having a hard time distinguishing the overflowing concepts: non-experimental design, observational research, correlational study. Question is simple: is there any difference between those three concepts?
If there is any, what is that precisely?
$\begingroup$
$\endgroup$
1
-
3$\begingroup$ Sounds to me like the first two may be essentially the same. The last one, "correlational study" sounds like a subset of the first two. A correlational study sounds like it's making zero claims of estimating causal effects, while, you can estimate causal effects from observational data in certain circumstances. Estimating causal effects from observational data is what econometrics is largely about! (Of course, how convincing the use of these methods are varies quite widely...) $\endgroup$– Matthew GunnCommented Oct 4, 2016 at 5:38
Add a comment
|
1 Answer
$\begingroup$
$\endgroup$
1
Partially answered in comments:
Sounds to me like the first two may be essentially the same. The last one, "correlational study" sounds like a subset of the first two. A correlational study sounds like it's making zero claims of estimating causal effects, while, you can estimate causal effects from observational data in certain circumstances. Estimating causal effects from observational data is what econometrics is largely about! (Of course, how convincing the use of these methods are varies quite widely ...)
– Matthew Gunn
answered Jan 18, 2019 at 12:49
-
$\begingroup$ A little Googling for "non-experimental design" indicates it is not equivalent to an observational study. Although it is difficult to find a clear definition of the former, the key concept seems to be "in a non-experimental design, the independent variable is not controlled." $\endgroup$– whuber ♦Commented Mar 3, 2023 at 20:22