Alba Regia. Annales Musei Stephani Regis. – Alba Regia. A Szent István Király Múzeum Évkönyve. 30. 2000 – Szent István Király Múzeum közleményei: C sorozat (2001)

Tanulmányok – Abhandlungen - Finnegan, M. –Éry Kinga: Biological distance among six population samples excavated in the environs around Székesfehérvár, Hungary, as derived by non-metric trait variation. p. 61–76.

breakage and pathology in the area of that trait, the sample size is variable for each trait. The actual number for each trait is recorded in Table 3. Note here that for some traits, 2N (twice the sample size due to the bilateral traits) is presented as ultimately left and right sides were pooled. Table 4 displays the minimum and maximum frequency and the range in frequency of each trait across the six population samples. In that we are interested in the biological distance between these population samples, we utilized FinneganCooprider (1978) for determining which statistic to use. We settled on the statistic developed by С A. B. Smith, which was first used by Grewal (1962). This is the statistic of choice in that it is easily computed and has been used in research concerning other regions of Hungary (FinneganMarcsik 1979, 1989a). RESULTS One of the advantages of the statistic used, and nonmetric trait analysis in general, is for bilateral traits information may be gathered from each side of the cranium. We tested for side asymmetry using the chi-square statistic based on theta values as presented in Finnegan (1972). This chi square test was used in analyzing side asymmetry in crania assessed as male for the 42 non-metric traits. Fourteen traits were significant at or above the .05 level. At this level of significance, we would expect, in this case, 10.5 significant differences due to chance alone. Here, we slightly exceed the chance expectation at the .05 level. While two traits were significantly different at the .01 level, this does not exceed the chance expectation of 2.1 traits. Three traits generated significant side asymmetry in more than one population (Pterion Form, Frontal Foramen Present, and Mylohyoid Grove Closed). The number of significant side asymmetries varied from one to three traits in each population. Of those traits, where a significant side asymmetry was generated, four instances showed the left side to have the higher incident frequency, while 10 traits showed the right side to have the higher incident frequency. On those traits where a significant difference was found in more than one population sample, two samples had the higher frequency on the left, while one other sample had the highest incident frequency on the right. In one case, Mylohyoid Groove Closed, the higher incident was on the right side in the two samples, which showed a significant side asymmetry. In each of these samples the sample size was sufficiently high to discount the significant difference due to small sample size. In general, it can be said that the significant differences generated in this study are relatively random among the population samples studied and in only a few instances specific to a particular trait under analysis. We feel that these borderline significant differences, as a whole, do not present a need to control for side asymmetry in the male sample. In the female sample, 14 traits were significant at or above the .05 level of confidence. Again, we would expect only 10.5 significant differences due to chance alone at the .05 level. Therefore, the female samples slightly exceeded the chance expectation at the .05 level of confidence. Two traits developed significant side asymmetry at the .01 level of confidence, but this does not exceed the number of significant differences at the .01 level due to chance expectation. As well, 7 of the 14 significant differences were generated where one or both of the sides tested had a sample size less than 12, which we believe adversely affects the chisquare statistic used here. The number of significant side asymmetries vary from 1 to 3 traits within a population sample. These were randomly distributed with respect to the traits in that the same trait did not show a significant difference in more than one population. As well, in seven of these significant side asymmetries, the higher frequency was seen on the left side and in seven instances, the higher frequency was seen on the right side. Of the 14 traits significant at or above the .05 level, 7 of these traits had sample sizes below 12 on one or both sides, which raises some doubt as to the validity of the significance generated for that particular trait. In general, the results for females, as with the males, are marginal with respect to significant side asymmetries. In that this is a preliminary descriptive report, we chose to overlook the few significant side asymmetries above those expected due to chance alone. We realize that in a final study of population divergence, some minor corrections would have to be made either by omitting certain traits or in generating samples where the number of sides are equal. The same chi square analysis was used to test for sex dimorphism, keeping left and right sides separate so that the minor side asymmetry (observed above) would not be reflected in the chi-square analysis for sex. The chi square values generated in comparing male and females on the left side only are presented in Table 5. Here we find 21 significant differences at or above the .05 level whereas we would expect, due to chance alone, only 16 significant differences. Therefore, at this level of significance, chance expectation was exceeded. As well, at the .01 level of significance, we found 5 differences where chance expectation alone would suggest but 2.5. While this discrepancy is real and some significant sex differences are found, we note that in seven traits where significant differences were generated, the sample size of one or both samples involved was equal or less than 15 which allows a consideration for the sample size being, in part, responsible for the generation of signifi62