You need to keep your sample sets of Skin Resistance and Skin Conductance distinct. Skin Conductance is the reciprocal of Skin Resistance and is measured in Siemens rather than Ohms.

I bought a skin conductance meter and have been playing around with it. It definitely needs to be re-balanced after each response or it will continue going haywire. 

What I have found in playing with it is that my skin conductance does not increase with sphincter contraction or with toe pressing, or biting my tongue or lips. It doesn't increase with doing complex mental math.   

However, it does increase with scary images (I imagine myself in a panic - being on board one of the 9/11 flights and knowing I was about to die) and that will increase conductance.

What really proves that there is nothing to looking at a GSR chart and diagnosing deception (the FDA won't allow GSR makers to advertise their product as capable of diagnosis absent stringent criteria) is that the single biggest response I get on the GSR is from breathing!  Inspiration causes an increase in conductivity and exhalation a decrease in conductivity. 

A therapist friend and I were playing with it (he doesn't believe in polygraph). The manner in which he asked questions also had a great impact on the GSR. When questions were asked in a forceful manner, conductance increased. When asked in a calm manner,  little response. When asked in a loud voice, increase and when asked in a quiet tone, little response. The questions asked are what would be relevant questions for me. Also, the words used in a question are very important. For example, if he used the word "sex," it got a reaction. If he used the word "consummate," it got little reaction. If he used the word "steal," it got a reaction, if he used the word "take" it did not get as much reaction. So the problem is that when people take a polygraph, the mere mention of certain words are enough to create arousal (not deception), and increase skin conductance to a level that might have you found as "deceptive."

Aluminum Chloride deodorant also makes the meter read lower as a constant, which means the differential between a spike and a baseline is much more muted. Aluminum Chloride, while it increases conductivity as opposed to potassium chloride due to the extra number of ions (4 to 2), actually inhibits the GSR because ACl acts on the eccrine sweat glands (fingers, palms, soles) - it absorbs into the dermis (which means if it is applied consistently a few days out - it does not "wash away" even after washing hands!. 

Anyway - these are just my observations with the meter - anyone else have a meter and anyone else have any experiences with it?

And for what it's worth - Xanax didn't change any readings, just made the readings come back down quicker after an increase in conductance. 

And the other problem with the GSR is that it gets warm, and can create sweat. Once sweat comes into pore and even if it is re-absorbed, it leaves the sodium behind, which in and of itself increases conductivity.