SUMMARY We examine auditor choice and audit fees in family firms using data from Standard & Poor's (S&P) 1500 firms. We find that, compared to non-family firms, family firms are less likely to hire top-tier auditors due to the less severe agency problems between owners and managers. Our results also show that family firms, on average, incur lower audit fees than non-family firms, which is driven by family firms' lower demand for external auditing services and auditors' perceived lower audit risk for family firms. Our additional analysis indicates that the tendency of family firms to hire non-top-tier auditors and to pay lower audit fees is stronger when family owners actively monitor their firms.

This study examines empirically the interaction effects of national culture and contextual factors (nature of the knowledge and the relationship between the knowledge sharer and recipient) on employees' tendency to share knowledge with co-workers. Quantitative and open-ended responses to two scenarios were collected from 142 managers (104 from the U.S. and 38 from the People's Republic of China). These two nations were selected due to their divergence on salient aspects of national culture, as well as their global political and economic importance. The focus on interaction effects was aimed at providing a more powerful test of culture's effects than simple comparisons of means typical of prior related research. Consistent with culture-based expectation, the quantitative results indicated that Chinese vs. U.S. nationals' openness of knowledge sharing was related to their different degrees of collectivism—the relative emphasis on self vs. collective interests—as well as whether knowledge sharing involved a conflict between self and collective interests. Also consistent with prediction, Chinese relative to U.S. nationals shared knowledge significantly less with a potential recipient who was not a member of their “ingroup.” Content analysis of the open-ended responses further showed that the quantitative results are the aggregated outcomes of trade-offs across cultural attributes and their interactions with contextual factors.

The goal of this study is to advance understanding of factors that may enhance or hinder knowledge sharing in public accounting firms and, in the end, provide practical recommendations for the firms. Attention to this topic is warranted for two reasons. First, today's regulatory environment and new auditing standards have broadened and intensified pressures on CPA firms to enhance the quality, effectiveness, and efficiency of the audit process. Second, knowledge and expertise are unevenly distributed among the members of the audit team. Thus, knowledge sharing can help CPA firms in leveraging the skills, knowledge, and best practices of their professional staff. Against this background, CPA firms' ability to effectively deploy knowledgesharing activities is increasingly vital to their competitive advantage, including gaining tangible benefits in terms of time and cost reductions. We draw upon prior research in accounting, organizational learning, psychology, and knowledge management to examine the role of three factors–information technology, formal and informal interactions among auditors, and reward systems–in encouraging knowledge sharing. We develop recommendations for public accounting firms and suggest several directions for future research.

The rapidly increasing availability of electronic health records (EHRs) from multiple heterogeneous sources has spearheaded the adoption of data-driven approaches for improved clinical research, decision making, prognosis, and patient management. Unfortunately, EHR data do not always directly and reliably map to medical concepts that clinical researchers need or use. Some recent studies have focused on EHR-derived phenotyping, which aims at mapping the EHR data to specific medical concepts; however, most of these approaches require labor intensive supervision from experienced clinical professionals. Furthermore, existing approaches are often disease-centric and specialized to the idiosyncrasies of the information technology and/or business practices of a single healthcare organization. In this paper, we propose Limestone, a nonnegative tensor factorization method to derive phenotype candidates with virtually no human supervision. Limestone represents the data source interactions naturally using tensors (a generalization of matrices). In particular, we investigate the interaction of diagnoses and medications among patients. The resulting tensor factors are reported as phenotype candidates that automatically reveal patient clusters on specific diagnoses and medications. Using the proposed method, multiple phenotypes can be identified simultaneously from data. We demonstrate the capability of Limestone on a cohort of 31,815 patient records from the Geisinger Health System. The dataset spans 7years of longitudinal patient records and was initially constructed for a heart failure onset prediction study. Our experiments demonstrate the robustness, stability, and the conciseness of Limestone-derived phenotypes. Our results show that using only 40 phenotypes, we can outperform the original 640 features (169 diagnosis categories and 471 medication types) to achieve an area under the receiver operator characteristic curve (AUC) of 0.720 (95% CI 0.715 to 0.725). Moreover, in consultation with a medical expert, we confirmed 82% of the top 50 candidates automatically extracted by Limestone are clinically meaningful.