Ellis W S, Starmer W T
Am J Hum Genet. 1978 Jul;30(4):366-76.
Törbel provides an interesting test case for the study of the relationship between inbreeding measured by pedigrees and inbreeding measured by isonymy. At the start of this investigation, we were aware that isonymy could introduce biases into the calculation of the inbreeding coefficient in either direction. However, it was expected that in Switzerland, inbreeding from isonymy would be an overestimate due to patrilocal residence and polyphyletic names. One way of dealing with this problem [13] was not to be concerned with the absolute value of inbreeding but only in the difference between estimates. Any bias introduced in the estimate itself disappears in such comparisons, so that a trend of inbreeding can be ascertained correctly. However, it was considered equally important to subject several populations to both a complete pedigree analysis and an isonymic analysis to determine the relationship between estimates of inbreeding. Despite the fact that several authors (Swedlund [18], for example) warned users of isonymy to exercise caution, the careless application of isonymy still persists. In the present study, estimates of inbreeding from isonymy were brought into line with other methods based on pedigree analysis and population size. However, it was possible to do this only in Törbel where pedigree depth was extensive and relatively complete. Similar corrections are possible only when the distribution of mono- and polyphyletic names is known and when migration data are reliable. If the trouble is taken to make these corrections, the same time and effort might as well be spent in pedigree analysis (when fairly complete ascertainment is possible) to achieve the same end result.
托尔贝尔为研究系谱测量的近亲繁殖与姓氏相同测量的近亲繁殖之间的关系提供了一个有趣的测试案例。在这项调查开始时，我们意识到姓氏相同可能会在两个方向上给近亲繁殖系数的计算带来偏差。然而，预计在瑞士，由于从夫居和多源姓氏，姓氏相同导致的近亲繁殖会被高估。处理这个问题的一种方法[13]是不关注近亲繁殖的绝对值，而只关注估计值之间的差异。估计本身引入的任何偏差在这种比较中都会消失，这样就可以正确地确定近亲繁殖的趋势。然而，对几个群体同时进行完整的系谱分析和姓氏相同分析以确定近亲繁殖估计值之间的关系也被认为同样重要。尽管有几位作者（例如斯韦德伦德[18]）警告姓氏相同法的使用者要谨慎，但姓氏相同法的不当应用仍然存在。在本研究中，基于姓氏相同的近亲繁殖估计值与其他基于系谱分析和人口规模的方法保持一致。然而，只有在托尔贝尔才能做到这一点，那里的系谱深度广泛且相对完整。只有当单源和多源姓氏的分布已知且迁移数据可靠时，类似的校正才有可能。如果不嫌麻烦进行这些校正，不妨将同样的时间和精力花在系谱分析上（当可以进行相当完整的确定时）以达到相同的最终结果。
