posted 05-31-2012 09:28 AM            

I've had it since it came out - as a beta tester. I would recommend it to anybody looking for a new instrument. Like the other manufacturers, the folks at Stoelting are very responsive. If the software doesn't do something you want it to do (within reason, of course), let them know and you'll soon see a change.

As far as data quality goes, I couldn't be more happy. I get good signals from most everybody. As a rule, I don't see those plunging EDAs you mentioned elsewhere. Some people have EDAs that tend to drift downward, but they're the exception and not the rule.

The beauty is that the CPS records the data and then redisplays each channel in its own "swim lane." That way, everything fits nicely on the screen. Any sensitivity adjustments you make while recording are only for your viewing pleasure. It doesn't change what's coming in or how it's redisplayed. You can zoom in, increase sensitivity, etc, for scoring purposes, but you never lose anything. The editing features are also nice. If you have a finger movement in the cardio, for example, you can ignore it to score it by hand, but the computer "sees" the increase as just that and considers it a large reaction. You can edit it out with the CPS (without altering the original data - it's all tracked and reversible), and then you can use the algorithm(s) to give you a second set of eyes.

(I know the thought of editing scares some people, but the Utah studies almost always edited as part of the process. It's usually not many cases, but it happens. Of course, if the manufacturer smooths or filters (a form of editing) behind the scenes, we're okay with that, but if the examiner does it....)

I think their most recent activity sensor is about as good as it gets. I like other instruments too - don't get me wrong - but I think Stoelting is onto something with the current activity sensor. I can almost see eyeblinks. If I have time, I'll see if I can post a comparison and show why I like it.

Their software has come a long ways since CPS-II, but I was happy with that system too. It wasn't the flashiest system, but it recorded good quality data. The PRO is the best of both worlds, all the bells and whistles and good data quality. (I'm no engineer. I can only speak from a user's perspective.)

posted 05-31-2012 10:26 AM            

It does not bother scientists much at all. Neither does some filtering as long as it is accounted for. Editing data can reduce the need for more complex statistics (with less power) in science when balanced designs are used. In polygraph it may mean fewer inconclusive results.

Normal protocol is to account for editing, interpolation, or replacement of data with average values, by calculating and reporting the proportion of data that was edited/interpolated/mean-replaced.

Less than 2% is generally no concern. Probably same with 5%. Between 5% and 10% is worth paying attention to the difference in scores that would result without editing/interpolation/mean-replacement. If more than 10% of of data is edited that would be a bit of a concern. These are, of course, arbitrary and blunt general guidelines.

.02

r

------------------
"Gentlemen, you can't fight in here. This is the war room."
--(Stanley Kubrick/Peter Sellers - Dr. Strangelove, 1964)