Z-score training

[Guest guest]

I'm very interested in hearing from any of you who are actually USING z-score training and how it works and what your results have been. I think I've been pretty clear over the years on this list about my personal bias re: training every brain toward some kind of " normal " --espeicially one, where I don't know what actually makes it up. I have always argued that the true goal of brain training ought to be to make each of us as fully OURSELVES as possible--not to make us more " normal " . The Z-score training approach obviously goes directly in the opposite direction. But I've recently heard some pretty astonishing claims being made for this (not necessarily by disinterested parties), and I'd like to learn more about it.

I know, of course, that the Z-score is the number of standard deviations a given reading is from the mean of a normative database for a specific client. In effect, all one is really doing in Z-score training is trading a threshold based on the client's actual brain activity for a " normed " value. As I understand it, the normative value against which we are measuring does not change. Every 6-year-old girl would have her 6 Hz activity with eyes closed trained to the same target. In order to achieve, say, a 70% scoring ratio, one would simply adjust the threshold number of Z-scores up or down instead of adjusting the threshold amplitude. I don't see the big difference.

I can certainly understand the benefit to a company like BrainMaster of offering Z-score training (another add-on package that users buy and perhaps selling more of their mini-Q machines), and I can understand the benefit to Bob Thatcher (selling more hits to his database), but I'm not sure I get what the actual benefit to the client or trainer might be. Anyone know if Roy or Lubar or Sterman or any of the other normative database purveyors are planning to make their databases available in this way?

My second big question about the z-score training has to do with how one chooses the measures to train. It would seem you would have to do a Q to see where the variables were that were most unstable. According to one article I read, there are 943 variables in the Thatcher database. If each of those has a Z-score calculated for it and only 5% have z-scores above 2, that's 47 measures of brain activity that need to be trained to " normalize " the brain. Which one(s) are the really important ones? Do you just start off with whatever happens to have the highest Z-score and train, say, the top 5 first? What is the desired effect this will have on the brain?

I am aware that Kirtley Thornton and have been doing work with LD clients training at task and training to change the coherence values for (generally) fast-wave activity between specific sites. The only way to know which points and what task is to get that info from the Q. Unfortunately, I don't believe Thatcher's database has Task data, so it's useless for that type of training. Instead, you would simply have a QEEG done on a client and look for the high z scores in coherences at task to tell you what to train. Then you could input them (not as Z-scores but) as simple upper and lower limits to the threshold, which you can do in any software and train to them.

Last question: BioExplorer has had for some time now, an object called the Range Threshold, though it doesn't cost extra or require additional equipment, and no one has gone out and marketed it heavily. It allows you to set thresholds in stndard deviations (the z-score minus the norms) for whatever measure you choose, but it sets them DYNAMICALLY (they are always updating based on performance), because it is comparing against the client's living brain, not against some fixed norm. You can set the target so the client scores only when within 2 standard deviations of their mean (more stable signal) or 1 standard deviation (even more stable) or 1.36 or whatever you want.

Standard deviation measures variability above and below the mean. So a client can be blocked from scoring when theta goes DOWN more than 2 standard deviations--not just when it goes UP! But the Range Threshold allows you to control the " offset " value, so you can set the target only to pick up movements in one direction, and allow the brain to go the other way as much as it wants. I haven't ever seen that the Z-score does this. Anyone know?

Looking forward to hearing from you.

Pete -- Van Deusenpvdtlc@...

http://www.brain-trainer.com305/433-3160The Learning Curve, Inc.