This paper argues that hospitality enterprises are a part of a complex-adaptive system. The exploration of Competitive Productivity (CP) and related concepts at the macro, meso, and micro (M-M-M) levels requires that we understand interrelationships beyond traditional analytical tools.

We propose three approaches to testing M-M-M CP architecture interactions: multi-level modeling, social network modeling, and agent-based modeling. Examples from the hospitality literature are offered for multi-level modeling and agent-based modeling, and we illustrate social networks by showing that IJHM authors network to create a highly competitive journal.

Multi-level modeling is useful for top-down testing (macro → meso → micro), but unsuitable for bottom-up modeling (micro → meso → macro) without making unrealistic assumptions about the nature of data. Interactions at the micro-level produce meso-level and macro-level behaviors, emergent phenomena that cannot be inferred by examining micro-level agents or behaviors. Social network theory and agent-based modeling are viable alternatives for examining and modeling emergent phenomena.

本文认为，酒店企业是复杂自适应系统的一部分。在宏观、中观和微观 (MMM) 层面探索竞争生产力 (CP) 和相关概念需要我们理解传统分析工具之外的相互关系。

我们提出了三种测试 MMM CP 架构交互的方法：多级建模、社交网络建模和基于代理的建模。酒店文献中的示例提供了多层次建模和基于代理的建模，我们通过展示IJHM的作者网络来创建一个竞争激烈的期刊来说明社交网络。

多级建模对于自上而下的测试（宏观→中观→微观）很有用，但不适用于自下而上的建模（微观→中观→宏观）而不对数据的性质做出不切实际的假设。微观层面的互动会产生中观层面和宏观层面的行为，这是无法通过检查微观层面的代理或行为来推断的突发现象。社会网络理论和基于代理的建模是检查和建模涌现现象的可行替代方案。