Abstract argumentation (AA), a non-monotonic reasoning framework, has attracted growing interest in the AI community. Explaining why an argument is accepted in AA emerges as a promising alternative for advancing explainable artificial intelligence (XAI). Driven by diverse practical needs, the logical structure and attributes of explanations must cater to varying contextual demands. This implies that multiple explanation methods are necessary to accommodate situational needs, with no single approach being absolutely superior or inferior. This paper categorizes the methodology of explanation in AA into five types: strong, presumptive, iterative, tree-like, and root explanation methods. To systematically capture the logical features underlying these explanation methods, we set twelve principles and ten properties, including but not limited to Non-Redundancy, Influence, and Composition. These principles and properties are employed to thoroughly analyze the five explanation methods. The results of this principle-and-property-based analysis highlight the importance of aligning explanation methods with practical requirements and provide valuable insights for future research in formal argumentation for explanation generation.