The validity of the Kinesiegraph (MHG-5R) output was studied using a non-ferromagnetic positioning device within working range of a 3 cm wide by 4 cm deep by 5 cm high three dimensional lattice. To determine how well observed values of the M K G outputs may predict case values of actual measurements, relationships between those values were tested by viewing scatter plots and correlation between observed values and actual values. In order to devise some form of served values of the M k G outputs (volts), statistical regression analysis was performed. The statistical analysis showed that observed values and actual values were strongly correlated. However, high correlation does not necessarily mean perfect predictability of given linear models. The formulated models were not able to predict all of acutal values this is partly 엳 새the discreteness of dependent variables (actual values) and also is because of independent nature of observed values. In conclusion, unless the Kinesiograph is suitably modified, the inherent nonlinearity characteristic of the system output limits its application to the measurement of a limited range of jaw movements.