I intend for this to be a very brief question. I am currently engaged in a small debate as to whether the study in this article is an observational study, or an experimental study?

Since initial effort and thought is always required here on CV, I vote that it is an observational study. My reasoning is that in an experimental study, the treatments are assigned to the experimental units where the researcher controls the levels of the treatments in an effort to measure effect(s) on the variable(s) of interest. I don't see that assignment happening here. Additionally, there are clear ethics implications in a study of this type, so an observational study is most likely going to be the only route to obtaining hopefully meaningful, if not limited, conclusions.

  • 3
    $\begingroup$ It is neither. Both experimental and observational study have some clear end result in mind inferred from the study. The article you have linked to is just a survey and it is not making any particular inference out of it. $\endgroup$ Commented Jun 12, 2017 at 6:32
  • $\begingroup$ This is definitely an aspect to think on further. I wish I could upvote your comment @TenaliRaman $\endgroup$
    – JMStats
    Commented Jun 14, 2017 at 4:19

1 Answer 1


It is no experiment. An experiment is defined by a) a randomised assignment to different groups. b) Each experimental group get a treatment that is controlled and systematically manipulated by the researcher.

Also, this survey doesn't have any hypothesis. I'm not sure whether this is needed but I would expect at least one hypothesis in a systematic study.

  • $\begingroup$ I'd say that an experiment does not require the assignment to be randomized. Also you do not require a control for an experiment. If I push some button and see what happens, that's already an experiment. You're talking about randomized controlled trials. $\endgroup$
    – ziggystar
    Commented Jun 12, 2017 at 7:10
  • $\begingroup$ Maybe this is domain specific, but in psychological research a true experiment has to have randomized assignment. If you don't randomly assign subjects to experimental groups you only get a quasi experiment. $\endgroup$
    – Gregor Ma
    Commented Jun 12, 2017 at 7:13

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.