In the title of your article you state "does not affect..." This is a strong assertion of no difference observed. That is, you claim to have demonstrated statistical support for the null hypothesis. I find this to be a faulty conclusion based on power issues indicated in previous queries and lack of appropriate testing methods.

You are trying to avoid a type II error. You are making a decision that there is no difference, and you want the chances of that decision being incorrect (there actually being a difference) to be very small. The probability of a type II error, Beta, is equal to 1-Power. Therefore, you must have large power to have small Beta.

Furthermore, there are statistical procedures for demonstrating equivalence, defined as an observed difference between conditions which has no practical or theoretical value (Rogers, Howard, & Vessey,1993 http://psycnet.apa.org/journals/bul/113/3/553/). These are not used in this study that I can see.

It seems that you hang a lot on the fact that there were observed differences in skin conductance, that this somehow shows the procedure to be sensitive. Procedures need to provide robust manipulations and that can be demonstrated by a manipulation check. It is the dependent variables that need to be sensitive and the sensitivity of one dependent variable does not usually imply much about the sensitivity of another dependent variable.