Firms increasingly use networks of alliances to pursue innovation. The current innovation literature has offered insights into how direct ties (between a focal firm and its partners, forming direct alliances) and indirect ties (between a focal firm's partner and its partners' partners but not including the focal firm, forming indirect alliances) function as independent antecedents to corporate innovation. It is, however, unclear how direct ties and indirect ties work in combination to impact innovation in a focal firm. Moreover, because different subtypes of financial or marketing alliances may operate with distinct governance structures and offer heterogeneous or incoherent resources for exchange, similarity in financial or marketing alliance subtypes, defined as the degree of overlap in financial or marketing alliance subtypes between direct and indirect ties, may significantly influence the extent to which corporate innovation can benefit from these ties. This study aims to examine the combined impact of direct and indirect ties on a focal firm's innovation by considering the moderating role of similarity in financial or marketing alliance subtypes. The results obtained by analyzing a longitudinal dataset extracted from US firms operating in the biotechnological and pharmaceutical industries support our hypotheses. Direct ties and indirect ties in combination have a negative effect on innovation as measured by patents and this effect is less negative when similarity in financial alliance subtypes is greater but more negative when similarity in marketing alliance subtypes is greater. We extend the innovation and alliance network literatures by offering novel evidence that direct and indirect ties in combination may diminish a focal firm's innovation and that such a negative combined effect depends on similarity in financial or marketing alliance subtypes.

Background: The purpose of this paper is to study how the Delta variant spread in a China city, and to what extent the non-pharmaceutical prevention measures of local government be effective by reviewing the contact network of COVID-19 cases in Xi’an, China. Methods: We organize the case reports of Shaanxi Health Commission into a database by text coding and convert them into structured network data. Then we construct a dynamic contact network for the corresponding analysis and calculate network indicators. we analyze the cases’ dynamic contact network structure and intervals between diagnosis time and isolation time by using data visualization, social network analysis method, and OLS regressions. Results: The contact network for this outbreak in Xi'an is very sparse, with a density of less than 0.0001. The contact network is a scale-free network. The average degree centrality is 0.741 and the average PageRank score is 0.0005. The network generated from a single source of infection contains 1,371 components. We construct three variables of intervals and analyze the trend of intervals during the outbreak. The mean interval (interval 1) between cases diagnosis time and isolation time is -3.9 days. The mean of the interval (interval 2) between the infector’s diagnosis time and the infectee diagnosis time is 4.2 days. The mean of the interval (interval 3) between infector isolation time and infectee isolation time is 2.9 days. Among the three intervals, only interval 1 has a significant positive correlation with degree centrality.Conclusions: By integrating COVID-19 case reports of a Chinese city, we can construct a contact network to analyze the dispersion of the outbreak. The network is a scale-free network with multiple hidden pathways that are not detected. The intervals of patients in this outbreak decreased compared to the beginning of the outbreak in 2020. City lockdown has a significant effect on the intervals that can affect patient’s network centrality. Our study highlights the value of case report text. By linking different reports, we can quickly analyze the spread of the epidemic in an urban area.