Does a confounding Variable always have an effect on the dependant variable?My textbook says that the confounding variable has an unwanted effect on the dependant variable, however i was under the...
Does a confounding Variable always have an effect on the dependant variable?
My textbook says that the confounding variable has an unwanted effect on the dependant variable, however i was under the impression that the problems with confounding variables was that they make you UNSURE of what caused the change in the Dependant variable, as i thought the problem was you could not PROVE it. I am unsure because if they have an unwanted effect, how can you also be confused, that would make the experiment not questionable but totally pointless. I am wondering if they always have an effect, or make things unclear?
First of all, the answer is that a confounding variable must have an effect on the dependent variable. Otherwise, it does not really matter, it does not confound the experimenter.
Let me point two things that I notice in your explanation here. First, you say that the presence of a confounding variable would make the experiment "not questionable but totally pointless." Those are really the same thing. If you cannot tell whether the effect on your dependent variable (DV) was caused by your independent variable (IV) your results are questionable and your experiment was pointless (except for the fact that it may hopefully have taught you to be sure to prevent extraneous variables from affecting the DV).
Second, you say the textbook says that the confounding variable (CV) has an unwanted effect on the DV. You go on to say that you are confused because you thought the problem was that the CV makes you unsure what caused the change to the DV. Both of these are absolutely correct. There should be no "however" there. The problem with having CVs is that they do cause an effect on the DV and you, as the researcher, are unsure whether the effect was caused by the IV or the CV. So this two statements are not contradictory. They go together perfectly.
So, a CV is only relevant if it has an effect on the DV. If some variable has no effect on the DV, we do not call it a CV. CVs are bad because they make us unsure as to whether the effect on a DV was caused by the IV or the CV, thus ruining our experiment.