ABSTRACT

In innovation literature, some view firms’ failures as negative to their innovative behaviour, while others argue that failures have a net function – learning from failure. However, we question whether classic linear regression methods constrain understandings of patterns by presenting only one side of the cause-and-effect relationship. This paper complements conventional approaches to literature by characterising the forms of innovation behaviour as mutually exclusive alternatives. Through an empirical approach evaluating mutually exclusive alternatives, we detected that firms’ failures simultaneously affect both the probability of choice to increase innovation investment and that to decrease innovation investment. Our results suggest that there is a certain threshold for the degree of failure to trigger innovative behaviour. When this is kept at a low level, innovative behaviour, rather than conservative behaviour, is more likely to be triggered. However, at a high level, companies could choose to increase their innovation investment; but they could also do the exact opposite at any time. Overall, attempts to model the innovation decision-making process as one on choosing the best out of mutually exclusive alternatives allows us to simultaneously compare the probability of innovation behaviour, elucidating the multidimensional aspects of innovation.

摘要

在创新文献中，一些人认为公司的失败对其创新行为不利，而另一些人则认为失败有一个净功能——从失败中学习。然而，我们质疑经典线性回归方法是否通过仅呈现因果关系的一侧来限制对模式的理解。本文通过将创新行为的形式描述为相互排斥的选择来补充传统的文献方法。通过评估相互排斥的备选方案的实证方法，我们发现公司的失败同时影响选择增加创新投资和减少创新投资的可能性。我们的结果表明，未能触发创新行为的程度存在一定的阈值。当这保持在较低水平时，更有可能触发创新行为，而不是保守行为。但是，在较高的层面上，企业可以选择增加创新投入；但他们也可以随时做完全相反的事情。总的来说，尝试将创新决策过程建模为从相互排斥的备选方案中选择最佳方案，使我们能够同时比较创新行为的概率，阐明创新的多维方面。