Class Predictions Research Articles

A cephalometric study of the accuracy of a computerized growth prediction in Class II four bicuspid extraction cases: Matt Mac Lean

A cephalometric study of the accuracy of a computerized growth prediction in Class II four bicuspid extraction cases: Matt Mac Lean

The use of electromagnetic measurements of apparent soil electrical conductivity to predict the boulder clay depth

Bulk soil electrical conductivity ( EC a) was measured under both dry and wet conditions by using the EM38 and EM31, in order to assess the potential of these instruments for measuring of the boulder clay depth ( D bc beneath coversand. The relationship between D bc and EC a was examined by fitting an exponential model (using D bc as a response variable) and a spherical model (using EC a as a response variable). The parameters of the exponential model were estimated by the maximum likelihood method, which makes it possible to take into account the censored observations of D bc. It was shown that the D bc predictions can be improved significantlly by stratification according to land-use. Nevertheless the residual standard deviation was rather large (43 to 67% of the predicted value), probably because the spatial variation in soil moisture content is not fully circumscribed by the stratification. Predicting D bc in two and four depth classes, gave only satisfactory results in the case of two classes: approximately 80% of the class predictions were correct. The results of the measurements of EC a, using the EM38 instrument, were better under wet conditions than under dry conditions. Additionally, under wet conditions, those with the EM38 were better than those with the EM31. Models with two measurements of EC a were not or only slightly better than models with a single measurement. The accuracy of the predictions with the spherical model was comparable to that of the exponential model.

The accuracy of computerized growth prediction in Class II high-angle cases

This study was conducted to test the accuracy of a commercially available forecasting system in predicting the effects of growth and orthodontic treatment. The pretreatment cephalograms and wax bites of mandibular casts of thirty-three consecutively treated Class II patients with high mandibular plane angles, along with twenty-six criteria related to treatment preference, were submitted for analysis. All patients had already been treated on a nonextraction basis by a single practitioner using high-pull face-bow headgear. The computer-generated posttreatment predictions or visual treatment objectives (VTO's) were compared to the actual posttreatment cephalograms, using twenty-one linear and nine angular measurements. Fifteen of the thirty parameters evaluated showed statistically significant (P < 0.01) differences between the actual posttreatment result and the computer prediction. The computer was found to be accurate in predicting the effects of growth and treatment on maxillary position and rotation, mandibular length, upper face height, and incisor positions. It was found to be inaccurate in predicting the effects of growth and treatment on maxillary length, mandibular rotation, lower anterior and posterior face heights, the horizontal and vertical positions of the molars, and over 50% of the soft-tissue parameters.

PREDICTION OF COLLEGE PERFORMANCE WITH THE MYERS‐BRIGGS TYPE INDICATOR1

ABSTRACTThis study, the last in a series by the present authors, investigated the ability of the Myers‐Briggs Type Indicator, a self‐report inventory, to predict grades and dropout at Wesleyan and Caltech, both for classes and, using estimates, for Caltech applicants. Continuous scores derived from the Indicator's four scales had some ability to predict the criteria, and this ability varied with the criterion and the sample. A contingency measure that reflects the Indicator's interdependent, dichotomous type categories generally had greater predictive validity than the continuous scores, but its correlations may be inflated. Adding the Indicator's continuous scores to the SAT scales and high school rank produced a slight but significant improvement in prediction in the Wesleyan class and the Caltech applicant group, but not in the Caltech class.