Talk:Design of quasi-experiments

I do not have time to edit this page but the links seem dead! 169.234.242.233 (talk) 17:55, 3 June 2008 (UTC)[reply]

Major Problems[edit]

The basic assumptions of this article are incorrect. A quasi-experiment is a correlational design, not an experimental design (despite the name). It can be used either in or out of the laboratory, as can true experiments. Although some quasi-experimental designs may provide greater external validity, this is NOT a feature of a quasi-experimental design per se. The only feature of quasi-experiments that separates it from both true experiment and correlational designs is that they are used for correltional studies when one variable is categorical, but can not be manipulated. All of the other features mentioned depend upon the particulars of the study and could be true of other types of research designs also.

Topics needed[edit]

I am just about to delete all but the opening from this version. I think the rest of the article is now mostly just a much longer and more confusing way of saying what the opening says clearly and succinctly.

The deleted text does contain pointers to subtopics that would be good to include, though. In no particular order, here's a list of topics that could make this an excellent article:

Interrupted and non-interrupted longitudinal studies
Pertinent statistical techniques
Important types of quasi-experiments
Controversy
Multiple meanings of the term
History
Some quasi-experiments that established important scientific results

-- Ben Kovitz (talk) 21:28, 9 February 2010 (UTC)[reply]

I have reverted the changes made above because what was left entirely omitted the topic of "design", while leaving content, and a potential content, which would be identical to what is, or might be, in the article natural experiment ....given that both the latter and the article Experiment say that these are identical. None of the proposed topics above seem to relate to "design". Thus there needs to be some thought about:

whether there should be an article on "design of quasi-experiments" or "design of natural experiments";
whether there can be a single article to cover both "quasi-experiment" and "natural experiment".

Notionally the latter question is already covered by the fact that quasi experiment already redirects to natural experiment, but is complicated by having quasi-experiments redirect to here (ie quasi-experimental design).

Melcombe (talk) 11:11, 11 February 2010 (UTC)[reply]

I don't understand what content was lost by the rewrite, aside from mentions of topics that I tried to include above. I do see what was lost: clarity. Instead of reverting, would you please tell more specifically what "design" would include, and that the rewrite left out? Please edit rather than reverting. -- Ben Kovitz (talk) 18:29, 11 February 2010 (UTC)[reply]

OK, I moved the rewrite to an article of its own, if you want to distinguish between an experiment and the design of an experiment. The present article is a mess, though. I look forward to seeing how you improve it. -- Ben Kovitz (talk) 19:02, 11 February 2010 (UTC)[reply]

It was clearly a case of necessary reversion since you had not only deleted all mention of "design", but also an obviously important citation. Since design of experiments presently explicitly excludes both natural experiments and quasi-experiments there is obviously a need for somewhere to say something about designing such things. For the record, I have now changed quasi experiment to redirect to quasi-experiment rather than natural experiment, as mentioned above. Melcombe (talk) 10:27, 12 February 2010 (UTC)[reply]

Merge?[edit]

Why do we have Quasi-experiment and Design of quasi-experiments? Should they be merged? Tayste (edits) 00:23, 9 August 2012 (UTC)[reply]