More often than not, when a phylogenetic   dataset is divided into smaller partitions, each one gives rise to trees that   have different topologies. One can draw two conclusions from this result: (1)   one, some or all of the trees are wrong and the partitions share the same   history, (2) or the trees are correct and the different partitions have   experienced distinct evolutionary histories. Distinguishing between these   options requires statistical testing to determine if the differences in   topology are likely to have been observed simply by chance.

There are many different tests of   incongruence available in the field of systematic biology that all use   comparable measurements and ideas. However, tests differ, sometimes subtly   and sometimes drastically, in their assumptions, implementation, and interpretation.   These details can be difficult to discern in the disjointed literature and   controversies surrounding these tests.

Incongruence tests may be broadly   classified into tests that consider character information (character   incongruence) and those that only consider tree shape or topology   (topological incongruence). Character congruence analyses are particularly   useful and powerful because they take both the tree topology and the   underlying support for the tree topology into account. Topological congruence   techniques have the advantage of being able to compare trees derived   from data that may not be strictly comparable or easy to include in the same   analysis.

通常情况下，当一个系统发育数据集被划分成更小的分区时，每个分区都会产生具有不同拓扑结构的树。从这个结果可以得出两个结论：（1）部分或全部的树是错误的，而各分区具有相同的历史；（2）或者树都是正确的，不同的分区经历了不同的进化历史。区分这些选项需要进行统计检验，以确定拓扑结构的差异是否可能只是偶然观察到的。

在系统生物学领域有许多不同的差异检验，都使用可比较的测量和想法。然而，检验在假设、实现和解释方面有时有微妙的，有时是显著的不同。在杂乱无章的文献和围绕这些检验的争论中，很难辨别出这些细节。

不一致性检验可以大致分为基础特征不一致的检验和只考虑树形状或拓扑（拓扑不一致）的检验。特征一致性分析特别有用和强大，因为它们同时考虑了树拓扑和对树拓扑的底层支持（数据来源）。拓扑同余技术的优点是能够比较来自数据的树，这些数据可能不具有严格的可比性，也不容易包含在同一分析中。