What are the differences between confounding and extraneous variables?
A confounding variable is one kind of extraneous variable.
In an experiment, you are trying to determine the impact that an independent variable has on a dependent variable. For example, you might try to measure the impact that two teaching strategies (independent variable) have on student performance on a test (dependent variable).
Extraneous variables are variables that are not the independent variable but which have an impact on the dependent variable. So, for example, if some students already know more than other students about the subject in the hypothetical experiment above, that would be an extraneous variable. Their prior knowledge (extraneous variable) would affect their test scores (dependent variable) regardless of the teaching strategy (independent variable) used.
Confounding variables are a kind of extraneous variable that is correlated with the independent variable. In other words, it changes along with the independent variable. This is very bad because it can make you think that your independent variable is having an impact on the dependent variable when really it's the extraneous variable that is having the impact. Going back to our example, lets say that all the students with prior knowledge get put in one group and all get the same teaching strategy. (Extraneous variable correlating to the independent variable.) That's going to be really bad because you'll think that teaching strategy was better when really all that happened was that those students had prior knowledge.
So, an extraneous variable is one that affects the dependent variable but is not the variable whose impact you're trying to study. A confounding variable is a very bad sort of extraneous variable that changes along with your independent variable and therefore makes you think your independent variable is having an impact when it really isn't.