臺大管理論叢 NTU Management Review VOL.29 NO.1

Choose Foreign R&D Partners From Right Pools: A Synthesis Framework 112 Foreign R&D partner selection . As aforementioned, we classify R&D partners into the internal stakeholders (ISTK) and the external stakeholders (ESTK). We code ISTK as 1 when the counter parties that the firms partner for R&D collaboration are customers, material suppliers, or subcontractors, otherwise as zero. We code ESTK as 1 when the counter parties are technical consulting firms, local professionals, technological institutes, or universities, otherwise as zero. Local institutional context . We use a dummy variable for institutional voids (IV) to examine how the local institutional context may moderate the partner selection decisions. We operationalized this variable by asking the firms what difficulties they have encountered when undertaking the outward FDI. We code IV as 1 when the respondents choose any of the following items, otherwise as zero. • Administrative inefficiency of local governments • Ambiguous regulations and complex local cliques • Insufficient local infrastructure • Complicated procedures of the customs of host countries • Difficulties in financing • Difficulties in liquidity • Risks in account receivables Control variables . We control two firm-specific characteristics, i.e., firm size and R&D intensity. Firm size (Size) is measured by the logarithm of the yearly revenue. R&D intensity (RD) is measured by the ratio of annual R&D expenditures to sales. Statistical analysis . We test our hypotheses by using path analysis with Stata/MP 13. The correlation matrix between variables (Table 2) reveals no multicollinearity among independent variables. The moderating effect of institutional voids (IV) on the relationship between the constructs of learning strategy and foreign R&D partner selection is examined with the interaction terms of the institutional voids and learning strategies.