[Examination of the degree of agreement among dental findings used for personal identification (report 5)--canonical correlation between dental findings on dental charts and those on radiographs].

Findings obtained from dental charts and from radiograph taken one year later for the same individuals were examined and classified according to different criteria. Similarities between the two sources investigated by the canonical correlation method. The results obtained were as follows: 1. The correlation coefficient and the squared multiple correlations (R2) were calculated. Consequently, high similarities were seen between the two sources for the mandibular anterior teeth [sequence: see text], maxillary incisor teeth [sequence: see text], mandibular and maxillary first premolar teeth [sequence: see text], and the corresponding teeth of the same jaws. 2. Three roots showed high elucidative values by the canonical correlation method. The first root was understood to be the mandibular incisor factor, the second the maxillary incisor factor, and the third the first premolar factor. This seemed to indicate, that the similarities between the two sources in these regions were very high. 3. A study of the similarities between the two sources was done based on the individual scores of canonical variable loadings calculated in this analysis. High similarities were seen at [sequence: see text] in category 1 (intact & C1) for chart findings and category 1'(intact, C1-C3 & resin filling) for radiographic findings: at [sequence: see text] in category 2 (C2-C4, inlay & filling of repair) for chart findings and category 3'(stump of tooth) for radiographic findings; and at [sequence: see text] in category 3 (complete crown, jacket crown & post crown) for chart findings and category 3'(stump of tooth) for radiographic findings. 4. The present results suggest that if proper case in taken, one set of information can serve as a substitute for the other. In personal identification, higher efficiency may be achieved if this is borne in mind.