Hi again,
That would depend on the mathematical procedure that you are using and the
experimental setting, I think (I typically do not do experiments but help
other people with experimental data to analyse the results, so I do not
have a definitve say on this). Without having a concrete example to talk
on, it is hard to decide. The variables that you are not aware of are part
of the error term in certain statistical procedures. There are other
procedures, that control for all the time-invariant within unit variance
(such as fixed effects) even if you are not actually measuring or aware of
them. Am I making this more confusing ? :) Or if you are randomising in
an experimental setting, you would be still controlling for that kind of
variables that you are not aware of but might be expecting the experiment.
Ali
On Thu, Mar 23, 2017 at 20:16 Luis A. Vasconcelos <[log in to unmask]> wrote:
> Hi ali,
>
>
>
> Thanks for your message! What if you are not controlling for them? What if,
>
> in fact, you are not aware of such variables?
>
>
>
> I know that what I'm looking for could be also framed as external or
>
> extraneous variables, but that doesn't tell much about whether such
>
> variables are a result of particular experimental procedures or maybe
>
> characteristics of the participants. I also came across 'situational
>
> variables', but these seem to refer more to the environment in which the
>
> experiments are conducted.
>
>
>
> I'm not sure!
>
>
>
> Best
>
-----------------------------------------------------------------
PhD-Design mailing list <[log in to unmask]>
Discussion of PhD studies and related research in Design
Subscribe or Unsubscribe at https://www.jiscmail.ac.uk/phd-design
-----------------------------------------------------------------
|