Shaner D J, Bamforth J S, Peterson A E, Beattie O B
Department of Anthropology, University of Alberta, Edmonton, Canada.
Am J Phys Anthropol. 1998 Aug;106(4):547-52. doi: 10.1002/(SICI)1096-8644(199808)106:4<547::AID-AJPA9>3.0.CO;2-F.
The primary goal of our study was to compare photogrammetric measurements with caliper-derived measurements. We also looked at the difference between caliper-derived measurements that were taken with and without the landmarks marked. Thirteen facial measurements were repeated ten times on two adult subjects as follows: 1) Calipers were used to take the measurements before the landmarks were marked on each subject's face; 2) the landmarks were then marked with a black pencil, and the calipers were used to take the measurements again; and 3) images were taken of each subject with the markings left on the face, and the measurements were extracted from these images. Compared with the caliper-derived data taken with the landmarks marked, the photogrammetric means and standard deviations were typically larger, leading us to conclude that there was a systematic difference between the data. The generally greater variation in the photogrammetric measurements was ascribed to poor conditions, such as shadows, oblique markings, and unmarked landmarks. When the data gathered by caliper with and without the landmarks marked were compared, a systematic difference was suggested by the number of statistically significant t-test probabilities. Marking the landmarks reduced the standard deviations in some measurements by controlling two sources of variation: differing pressure on the skin and slippage of the calipers. Anthropologists, medical geneticists, and others who use measurements for diagnostic or classificatory purposes should be aware that data gathered by different techniques may yield different results.
我们研究的主要目的是比较摄影测量法与游标卡尺测量法的结果。我们还研究了标记地标和未标记地标时游标卡尺测量结果的差异。在两名成年受试者身上对13项面部测量进行了10次重复测量，具体如下：1）在每个受试者面部标记地标之前，使用游标卡尺进行测量；2）然后用黑色铅笔标记地标，再使用游标卡尺进行测量；3）拍摄每个受试者面部带有标记的图像，并从这些图像中提取测量数据。与标记地标后游标卡尺测量得到的数据相比，摄影测量法的平均值和标准差通常更大，这使我们得出结论，两种数据之间存在系统差异。摄影测量中普遍较大的变异性归因于不良条件，如阴影、倾斜标记和未标记的地标。比较有标记地标和无标记地标时游标卡尺收集的数据，统计学上显著的t检验概率数量表明存在系统差异。标记地标通过控制两个变异来源——皮肤上不同的压力和游标卡尺的滑动，降低了某些测量的标准差。使用测量数据进行诊断或分类的人类学家、医学遗传学家及其他人员应意识到，不同技术收集的数据可能会产生不同的结果。
