Collaborative Expedition Workshop #71 National Science Foundation Technology Program Evaluation: Methodologies from the Advanced Technology Program Research Applications of Data Andrew Wang Technology Innovation Program National Institute of Standards and Technology Collaborative Expedition Workshop #71 National Science Foundation March 18, 2008
Research Applications Using Data Use of patents as a measure of R&D output Use of survey data to assess R&D alliances
Use of Patents Patents as a measure of R&D output What do patents measure? Number of patents Quality of patents
Patent Data and Measures Patent data sources Administrative data Survey data Archival data Constructing patent measures Project-level Firm-level Industry and Technology field
Research Application: ATP Project Structure and Patent Outcomes Joint Ventures University participation Patent data Firm-level, 1988-1999 ATP projects, 1990-1995 Firm participation vs. Firm-project participation Trends in propensity to patent
Research Application: ATP Project Structure and Patent Outcomes Regression analysis Panel data, 1988-1996 Before and after comparison, treatment effect Patents, by date of application, one year lag Control variables: firm size, technology field Patent outcomes ATP project increases firm patenting 4%-25% Joint Venture and University participation have positive effect on patenting
Use of Survey Data Survey measures What do surveys measure? Quantitative data, objective, “hard” data Qualitative data, subjective, perceptual
Research Application: Determinants of Success in R&D Alliances What factors influence the success of R&D alliances in generating innovation? Alliance Formation factors Alliance Structure (e.g., number, type of partners) Partner attributes (e.g., prior experience, capabilities) Alliance Execution factors Commitment and Communication Governance and Trust
Research Application: Determinants of Success in R&D Alliances What are measures of success of R&D alliances in generating innovation? Qualitative assessment of project success and value Achieving technical objectives Intangible value Patent applications Commercial impact (e.g., cost savings, revenues)
Research Application: Determinants of Success in R&D Alliances Survey of ATP Joint Ventures Data for 397 firms in 142 R&D alliances Findings from regression analysis Alliance execution factors are better predictors of success than alliance formation factors Commitment of human capital and frequency of communication are key factors in success Contractual trust matters, but goodwill trust does not More ambitious projects are more likely to be successful