The New Mind matrix is cross validated against Nx Link. Neuroguide, and the Sterman Kaiser Databases with respect to normative dimensions.  We use our own collected data to statistically analyze the distribution of the clinical population as well (we have presently over 10,000 maps with associated psychometric, biometric, and sociometric  data).  There are good statistical reasons for wanting to know the variance in a presumed “normative” population and the variance in a “clinical” population (see for example  Romano-Micha, J. (2003, Nov). “Databases of Specific Training Protocols for Neurotherapy? A Proposal for a ‘Clinical Approach to Neurotherapy’”. Journal of Neurotherapy, 7 (3/4), pp 69-85.).  Using this information, our clinical significance levels are adjusted to correlate most accurately with psychometrics and symptomology.  We have used measures such as the Beck Inventories, the TOVA, and Microcog  as references as well as our own internally generated instruments and checklists.  This data is statistically analyzed and reviewed on a regular basis.  We believe this provides a unique form of “Reference” database anchored in the profoundly important work in normative neurometric research done by E Roy John, Bob Thatcher and David Cantor.  I think comparing what we are doing with normative databases as comparing apples to oranges.  They are different neurometric tools for different purposes.  Our purpose is to provide a high quality database that is very clinically friendly, simple to use and clinically accurate for purposes of doing neurofeedback.  I developed this database because I was frustrated with the features of the other databases I used and I wanted something different for clinical work.  Others who experience similar frustrations tend to use our system as well.  We are not interested in developing a system for research, medical applications or diagnoses.  Other databases in the field appear more interested in providing those features and services and some provide very sophisticated and complex features for these purposes.   I recommend those interested in doing advanced analysis and neurometric research use Neuroguide or Brain DX.

 

We are presently further validating our approach to neurometric analysis through research in collaboration with the University of North Carolina and Brainmaster and we are looking at covariance in neurometric and psychometric data with respect to academic performance.  Here we hope to further clarify the difference between normal, abnormal and peak performance features of qEEG. 

 

The checklist in our system is based on an extensive review of the fMRI literature as well as the neurology literature in order to predict which locations in the 10-20 system are most likely to show abnormal variance as a consequence of endorsed items.  This generates a predictive map which is then compared with the acquired EEG map for correlations.  The outcomes are understood as probabilities with the expectation that there will be many false positive and negatives because of the profound complexity of the brain and the dynamics of associated network functions.  Agreement between client endorsements and EEG abnormalities result in increased confidence that activities in these locations are associated.  In this sense we are looking at the relationship between a symptom and an array of locations.  We are not looking at the relation between qEEG and a diagnostic category.  I believe this is a dead model that does not fit with the neurofeedback paradigms. 

 

The protocols generated in our database system are not based on client endorsements of the items in our checklists (which I think you are suggesting?), they are instead generated by nonparametric  rank order analysis of the deviations observed at each location in each standard dimension of neurometric analysis in order to determine the most deviant locations.  We then utilize a combination of protocols reported effective in the peer reviewed literature(largely from the Journal of Neurotherapy) and 20 years of well documented clinical experience to generate probable protocol solutions that appear to best fit the deviant distribution features in the map.  Those wishing to better understand what we are doing should take our brief course on qEEG at www.newmindacademy.com  and attend our lunch and learns as well as experiment with our map system. Those interested in more detailed documentation regarding our database system are invited to review the New Mind Maps website.