Szekessy Vilmos (szerk.): A Magyar Természettudományi Múzeum évkönyve 60. (Budapest 1968)
Bottyán, O.: An analysis of the palatal measuring methods
Correlations were established also for the Hungarian sample. The correlational values of lengths, found by the two kinds of measuring methods are r= 0.77 for the the males, and r = 0.76 for the females. In the case of the breadths, the correlational values are r = 0.45 for the males, and r = 0.42 for the females (cf. Graphs 1—4 and 9-12). The correlational values between the lengths and breadths, by the classical measuring method, are r = 0.23 for the males, and r = 0.36 for the females. By the HEINTZ — method, r = 0.25 for the males, and r = 0.32. for the females (cf. Graphs 5-8 and 13-16). The estimated values of the probability levels pertaining to these correlational coefficients are as follows : the correlations of the relationships 1 and 2 with respect to Table 1 are so high that their probability levels cannot be computed by simple calculation methods. The probability levels between the correlations (female and male) of the breadth mesurements (No. 2) stand near to the probability level of the longitudinal correlations (No. 1) of HEINTZ'S sample. The probability levels pertaining to the correlations (No. 1) between the longitudinal measuremets of the sample from Hungary are manifestly more restricted than the preceding ones. The probability levels, estimated on the basis of the correlations between the longitudinal and transversal measurements, are, by the classical measurements (No. 3), 10~ 3 in the males and 10~ 5 in the females, whilst by the HEiNTZ-measurements (No. 4) they are 10~ 4 in the males and 10-6 in the females. It follows that significant correlations were found in both sexes with respect to all four relationships for the Hungarian sample. Comparison of results As regards the rate of correlation, considerable deviations were established between HEINTZ'S and the Hungarian samples, (cf. Table 1). In the order of importance, they are as follows : 1. By HEINTZ'S sample, the classical longitudinal and transversal measurements (No. 3) do not correlate. By the Hungarian sample, these measurements correlate significantly. I have established with respect to the differences of these two results that the deviation concerning the female data of the Hungarian sample is significant, its probability about 97 per cent. For the male data, the probability of deviation is about 86 per cent. 2. By HEINTZ'S sample, the breadth measurements taken by the classical and the HEiNTZ-method (No. 2) do not correlate. By the Hungarian sample, these measurements correlate significantly. The deviation between these two results, with respect to the male data of the Hungarian sample, is significant, its probability about 95 per cent. Concerning the female data, the probability of deviation is about 92 per cent. 3. By HEINTZ'S sample, the correlations between the longitudinal measurements taken by HEINTZ'S and the classical methods (No. 1), as well as those between the longitudinal and transversal measurements taken by HEINTZ'S method (No. 4), are smaller than by the Hungarian sample. In the latter case (No. 4), the probability levels by HEINTZ'S sample are about 10-2 , by the Hungarian sample about 10~ 5 . In the former case (No. 1), deviation between the two probability levels is of the same rate and, naturally, of the same direction. On the basis of the results deriving from the examination of her sample, HEINTZ made the following two statements : 1. With respect to the relationship between the
