Procedures for Computing the Mean Age of Eruption of Human Teeth

SEVERAL different methods have been used for computing the mean ages of eruption of the teeth from eross-sectional population data. The earlier methods have been reviewed by other authors."-` Some of the methods, such as those used by Cattell' and Cohen,3 are primarily graphical procedures which require no assumption about the underlying distribution of the ages of eruption. These procedures yield good descriptions of tooth eruption for the group being studied but they do not permit judging the significance of differences among tooth types or among populations. Other investigators have suggested that the ages of eruption of a tooth conform to the normal probability curve.2' 4, 5 The procedures used by Boas' and Klein2 and their co-workers yield estimates of the mean and standard deviation of the ages of eruption but neither of these authors discusses technics for making statistical comparisons between populations. Clements, Davies-Thomas, and Pickett5 applied a more complicated statistical procedure which includes provisions for computing the confidence limits of the mean. This latter procedure, then, provides for estimating the significance of differences in mean ages of eruption, but the computations are so arduous that it may not be suitable for routine use by dental investigators. Both Klein2 and Clements' found that the distribution of the ages of eruption was satisfactorily described by the normal probability curve. Another team of investigators, Kihlberg and Koski,6 suggested the use of the logarithms of the ages of eruption, with conception as the zero point of age. These investigators found that these logarithms distribute normally and that the logarithmic standard deviations of the different tooth types have similar magnitudes. This suggestion that the logarithms of the ages of eruption distribute normally differs from the findings of other investigators.2 The present report introduces a simpler procedure for computing eruption statistics and compares the results with those obtained by more complex procedures. It also attempts to test, for a selected tooth type with a relatively uncomplicated eruption pattern, the assumptions that (a) the ages of eruption are normally distributed, and (b) the logarithms of the ages of eruption are normuallv distributed.