Appraising the construct validity of product innovation capability and identifying its association with firm performance through meta-analysis

In the literature focused on dynamic capability and innovation constructs, a prominent theme is the identification and analysis of antecedents that drive firm performance. In particular, the dynamic capability construct of product innovation capability (PIC), and its implications for firm performance, has spurred substantial scholarly interest. This has eventuated into a considerable amount of theoretical and empirical literature on the PIC and firm performance relationship. The accumulation of empirical evidence on the PIC‚Äö-firm performance relationship has attained a critical mass that warrants and enables a systematic synthesis of findings. Notwithstanding the advancements made in understanding PIC, and its relationship with firm performance, several gaps and contradictions persist in the literature. For example, empirical findings concerning the relationship are often mixed, and theoretical contentions of Dynamic Capability (DC) Theory in general, have sometimes remained empirically unsubstantiated. The present study aims to advance DC Theory and innovation literatures by: 1. Undertaking a review of PIC measurement using theoretical triangulation, that entails a multi-theoretical appraisal of PIC construct validity. 2. Formulating an innovative meta-analytic methodology and conducting an investigation of the PIC‚Äö-firm performance relationship and its moderation effects via a statistical synthesis of findings. In undertaking a meta-analysis of the relationship between PIC and firm performance, the thesis firstly focuses on operationalisation of PIC in order to assess its construct validity as a dynamic capability construct. High construct validity of PIC is a necessary condition for enabling the development, empirical testing and application of DC Theory. Attaining high validity of PIC is also imperative for an assessment of its association with firm performance. Since DC Theory is arguably still nascent, particularly in terms of its scant empirical validation, PIC construct validity assessment serves to consolidate the theoretical and empirical underpinnings of the theory. A critical gap is identified in PIC measurement models and a novel meta-analytic methodology for addressing the validity problem is developed and employed in the study. The meta-analysis of the PIC‚Äö-firm performance relationship aggregates 81 effect sizes (correlations), extracted from 57 studies, representing the magnitude and direction of this relationship. The synthesis enables the computation of a summary (cumulative) effect size for the relationship under investigation. The synthesis also offers insights into certain boundary conditions, under which the magnitude and/or direction of the focal relationship undergo a change. This is accomplished through a priori identification and sub-group analyses of potential moderator variables. By ascertaining the moderation effects concerning the PIC‚Äö-firm performance relationship, DC Theory can also be better understood. The meta-analytic results demonstrate a positive and strong association between PIC and firm performance, supporting the hypothesised relationship and yielding a point estimate for the true (i.e., construct-level) relationship. In other words, the current study provides a summary estimate of the actual underlying bivariate relationship of interest, by overcoming an identified construct validity problem that limits the existing PIC operationalisation methods. The validity problem in PIC, as determined through triangulation and relevant arguments, and the development of a unique meta-analytic model in this study, provide a broad spectrum of opportunities for further research.