TY - JOUR A1 - Sirianni, Antonio D. A1 - Morgan, Jonathan H. A1 - Zöller, Nikolas A1 - Rogers, Kimberly B. A1 - Schröder, Tobias T1 - Complements and competitors BT - Examining technological co-diffusion and relatedness on a collaborative coding platform JF - PNAS nexus N2 - Diffusive and contagious processes spread in the context of one another in connected populations. Diffusions may be more likely to pass through portions of a network where compatible diffusions are already present. We examine this by incorporating the concept of “relatedness” from the economic complexity literature into a network co-diffusion model. Building on the “product space” concept used in this work, we consider technologies themselves as nodes in “product networks,” where edges define relationships between products. Specifically, coding languages on GitHub, an online platform for collaborative coding, are considered. From rates of language co-occurrence in coding projects, we calculate rates of functional cohesion and functional equivalence for each pair of languages. From rates of how individuals adopt and abandon coding languages over time, we calculate measures of complementary diffusion and substitutive diffusion for each pair of languages relative to one another. Consistent with the principle of relatedness, network regression techniques (MR-QAP) reveal strong evidence that functional cohesion positively predicts complementary diffusion. We also find limited evidence that functional equivalence predicts substitutive (competitive) diffusion. Results support the broader finding that functional dependencies between diffusive processes will dictate how said processes spread relative to one another across a population of potential adopters. KW - Computational social science KW - Innovation KW - Netzwerk Y1 - 2024 U6 - https://doi.org/10.1093/pnasnexus/pgae549 SN - 2752-6542 VL - 3 IS - 12 SP - 1 EP - 13 PB - Oxford University Press CY - Oxford ER - TY - JOUR A1 - Smith, Jeffrey A. A1 - Morgan, Jonathan Howard A1 - Moody, James T1 - Network sampling coverage III BT - Imputation of missing network data under different network and missing data conditions JF - Social Networks N2 - Missing data is a common, difficult problem for network studies. Unfortunately, there are few clear guidelines about what a researcher should do when faced with incomplete information. We take up this problem in the third paper of a three-paper series on missing network data. Here, we compare the performance of different imputation methods across a wide range of circumstances characterized in terms of measures, networks and missing data types. We consider a number of imputation methods, going from simple imputation to more complex model-based approaches. Overall, we find that listwise deletion is almost always the worst option, while choosing the best strategy can be difficult, as it depends on the type of missing data, the type of network and the measure of interest. We end the paper by offering a set of practical outputs that researchers can use to identify the best imputation choice for their particular research setting. KW - Missing data KW - Imputation KW - Network sampling KW - Network bias KW - Fehlende Daten KW - Imputationstechnik Y1 - 2022 U6 - https://doi.org/10.1016/j.socnet.2021.05.002 SN - 0378-8733 VL - 68 SP - 148 EP - 178 PB - Elsevier CY - Amsterdam ER -