An official website of the United States government
Although much attention is accorded to star performers, this paper considers the extent to which stars, themselves, benefit from the contribution of their collaborators (the constellation). By considering stars, constellations, and the synergies between them, we address a key question: To what extent is collaboration performance driven by the great individual or by great constellations? We introduce a novel approach that uses a matching model to uncover the complementarities driving collaboration formation. We use formal value-capture theory to estimate the relative contribution of stars and constellations to joint value creation. Analyzing a sample of academic research collaborations, we document that stars’ relative contribution exceeds that of their constellations in less than 15% of collaborations, although constellations provide a greater relative contribution in 9%. In most collaborations, neither party dominates: Innovation is a collective endeavor driven equally by the star and the constellation. Joint value creation and relative contribution are explained by the subtle interplay between complementarities in joint work and the substitutability of collaborative parties in the market. Joint value creation increases with the strength of complementarities between parties in a match. Relative value creation, and hence dominance, increases with the substitutability of one’s collaborative partner. Interestingly, joint value creation is greatest in collaborations where both stars and constellations offer bundles of rare attributes and where neither the star nor the constellation dominates. This paper was accepted by Olav Sorenson, organizations. Funding: D. Mindruta gratefully acknowledges funding from the HEC Foundation and from the French National Research Agency (ANR) “Investissements d’Avenir” (LabEx Ecodec/ANR-11-LABX-0047). J. Bercovitz and M. Feldman gratefully acknowledge funding from the Science of Science Approach to Analyzing and Innovating the Biomedical Research Enterprise (SCISIPBIO) program of the U.S. National Science Foundation (NSF) [Grant 1934875]. Supplemental Material: The online appendices and data files are available at https://doi.org/10.1287/mnsc.2021.01969 .
more »
« less
With Great Training Comes Great Vulnerability: Practical Attacks against Transfer Learning
- Award ID(s):
- 1834523
- PAR ID:
- 10079885
- Date Published:
- Journal Name:
- Proceedings of the 27th USENIX Security Symposium
- Format(s):
- Medium: X
- Sponsoring Org:
- National Science Foundation
More Like this
-
-
The questions we ask and how we ask them will make a difference in how successful we are in meetings, in collaborations and in our careers as statisticians and data scientists. What makes a question good and what makes a good question great? Great questions elicit information useful for accomplishing the tasks of a project and strengthen the statistician–domain expert relationship. Great questions have three parts: the question, the answer and the paraphrasing of the answer to create shared understanding. We discuss three strategies for asking great questions: preface questions with statements about the intent behind asking the question; follow the question with behaviours and actions consistent with the prefaced words including actions such as listening, paraphrasing and summarizing; and model a collaborative relationship via the asking of a great question. We describe the methods and results of a study that shows how questions can be assessed, that statisticians can learn to ask great questions and that those who have learned this skill consider it to be valuable for their careers. We provide practical guidelines for learning how to ask great questions so that statisticians can improve their collaboration skills and thus increase their impact to help address societal challenges.more » « less
-
Great Power Brings Great Responsibility: Personalizing Conversational AI for Diverse Problem-SolversNewcomers onboarding to Open Source Software (OSS) projects face many challenges. Large Language Models (LLMs), like ChatGPT, have emerged as potential resources for answering questions and providing guidance, with many developers now turning to ChatGPT over traditional Q&A sites like Stack Overflow. Nonetheless, LLMs may carry biases in presenting information, which can be especially impactful for newcomers whose problem-solving styles may not be broadly represented. This raises important questions about the accessibility of AI-driven support for newcomers to OSS projects. This vision paper outlines the potential of adapting AI responses to various problem-solving styles to avoid privileging a particular subgroup. We discuss the potential of AI persona-based prompt engineering as a strategy for interacting with AI. This study invites further research to refine AI-based tools to better support contributions to OSS projects.more » « less
- Free Publicly Accessible Full Text
-
Accepted Manuscript1.0
- Conference Paper:
- The DOI is not currently available.
