Comparing automated text classification methods

Online social media drive the growth of unstructured text data. Many marketing applications require structuring this data at scales non-accessible to human coding, e.g., to detect communication shifts in sentiment or other researcher-defined content categories. Several methods have been proposed to automatically classify unstructured text. This paper compares the performance of ten such approaches (five lexicon-based, five machine learning algorithms) across 41 social media datasets covering major social media platforms, various sample sizes, and languages. So far, marketing research relies predominantly on support vector machines (SVM) and Linguistic Inquiry and Word Count (LIWC). Across all tasks we study, either random forest (RF) or naive Bayes (NB) performs best in terms of correctly uncovering human intuition. In particular, RF exhibits consistently high performance for three-class sentiment, NB for small samples sizes. SVM never outperform the remaining methods. All lexicon-based approaches, LIWC in particular, perform poorly compared with machine learning. In some applications, accuracies only slightly exceed chance. Since additional considerations of text classification choice are also in favor of NB and RF, our results suggest that marketing research can benefit from considering these alternatives.     

Expected earnings growth and the cost of capital: an analysis of accounting regime change in the European financial market

This study focuses on the relation between the cost of equity capital and earnings expectations when the properties of accounting that determine earnings vary across different regulatory regimes. More particularly, it addresses the European setting where different types of GAAP regime have continued to function in the presence of the gradual harmonization of the underlying legal framework, and where the adoption of internationally recognized accounting standards by certain firms has anticipated the requirement for International Financial Reporting Standards. On the basis of estimates of the cost of equity that are implied by analysts' earnings forecasts, the article provides evidence that financial market integration may have already contributed to mitigating the economic consequences of accounting diversity, and that switching to IFRS could have a short lived impact on capital markets. Moreover, based on firm level transparency and disclosure rankings provided by Standard and Poor's, it is shown how the quality of financial reporting conditions the implied cost of equity under different GAAP.     

International Framework Agreements and Industrial Relations Governance: Global Rhetoric versus Local Realities

The article discusses the local impact of international framework agreements (IFAs). It presents findings from empirical research carried out in two Spanish MNCs, and it examines how the commitments stipulated by the texts of the IFAs translate in local practice in different subsidiaries in Central and South America. The article follows the policy cycle of both agreements, and it further evaluates them as emerging private institutions of global industrial relations governance. The main findings reveal a paradox. The generic character of IFAs encourages corporate consent for their adoption, but, in practice, it gives precedence to local laws and hence renders the enforcement of compliance difficult, and, in some cases, unfeasible. And where local laws contravene or lack compatibility with International Labour Organization conventions, the central premise of IFAs as instruments of global governance is largely undermined.     

Trust as a mediator of the relationship between leader/member behavior and leader-member-exchange quality

We examined how and why trust between a subordinate and supervisor develops over time from the separate perspectives and experiences of both dyad partners. Integrating the literatures on trust, leader-member-exchange (LMX) and promotion regulatory focus (PRF), we test our model using longitudinal data from a Malaysian subsidiary of a U.S. headquartered corporation. Subordinate's trust in supervisor mediated between supervisor's PRF and subordinate's experience of LMX quality; supervisor's trust in subordinate mediated between subordinate's job performance and leader's experience of LMX quality. Findings are interpreted within the high power-distance, high collectivist, and high performance orientation of Malaysian culture.