Pith. sign in

REVIEW 3 major objections 62 references

Share-feature A/B tests can recover the platform-wide flywheel effect with a closed-form adjustment from attribution logs, not cluster randomization.

Reviewed by Pith at T0; open to challenge. T0 means a machine referee read the full paper against a public rubric. the ladder, T0–T4 →

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗

T0 review · grok-4.5

2026-07-14 09:00 UTC pith:SYC4DF7F

load-bearing objection Clean, deployable GTE estimator for sharing flywheels; A2 is a real but already-stressed soft spot, not a collapse. the 3 major comments →

arxiv 2607.10820 v1 pith:SYC4DF7F submitted 2026-07-12 stat.ME

Causal Estimation of Share-Induced Engagement with Flywheel Effects

Weitao Cheng , Yilin Li , Yong Wang , Nian Si This is my paper
classification stat.ME
keywords A/B testsexperimental designcausal inferenceinterferencesocial networkglobal treatment effectHawkes processflywheel effect
verification ladder T0 review T1 audit T2 compute T3 formal T4 reserved

The pith

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

Online platforms use sharing features to reactivate users and amplify engagement through multi-round social cascades, the so-called flywheel effect. Classical A/B tests miss this impact because one user’s treatment changes what others see and share, violating the no-interference assumption. This paper models sharing as a multivariate Hawkes process and shows that share-induced engagement obeys a sender–receiver flow-balance identity. From that identity it builds a plug-in estimator that treats multi-round diffusion as geometric amplification and corrects for it using ordinary attribution logs. Under mild conditions the estimator is consistent for the global treatment effect, an A/A procedure controls Type I error, and simulations plus a large-platform deployment show lower bias than difference-in-means or first-order adjustments—sometimes flipping an insignificant launch decision into a significant one.

Core claim

The global treatment effect of a sharing feature on share-induced engagement equals the difference, between full treatment and full control, of discovery-driven share-view volume scaled by the geometric multiplier 1/(1−q_eff), where q_eff is the effective downstream sharing rate. This quantity is identified from Bernoulli A/B data by the plug-in estimator that replaces discovery-driven offspring counts and the two group-level rates with their sample analogues from attribution logs, and the estimator is consistent when downstream rates are homogeneous within regime.

What carries the argument

The flow-balance identity equating total sender-side offspring share-views to total receiver-side share-induced views; it yields the geometric representation IS = (average discovery-driven offspring)/(1−q_eff) and the closed-form propagation-adjusted GTE estimator.

Load-bearing premise

Downstream sharing rates are the same for every user–content pair within a treatment regime; if those rates systematically differ by who is treated or who receives the content, the single geometric factor is wrong and bias returns.

What would settle it

On a platform where per-user or per-content downstream share rates can be measured separately under global treatment and global control, check whether the single-rate plug-in GTE matches the true full-deployment difference; large systematic gaps would falsify the homogeneous-rate claim that carries consistency.

Watch this falsifier. Get emailed when new claim-graph text bears on it.

If this is right

  • Product teams can evaluate multi-round sharing features with ordinary Bernoulli randomization and existing attribution logs, without redesigning experiments as graph clusters.
  • Difference-in-means and first-order cascade metrics systematically understate flywheel impact and can miss launch-worthy effects that the propagation-adjusted estimator detects.
  • A valid A/A pipeline check exists: under identical treatment and control the estimator is asymptotically normal with a plug-in standard error that keeps Type I error near nominal.
  • The same discovery-driven volume and downstream rates give a Poisson-based estimator for the global effect on user reactivation probability.
  • When the estimator is significant and baselines are not, platforms have a quantitative basis to launch and to validate post-launch against pre-launch metrics.

Where Pith is reading between the lines

These are editorial extensions of the paper, not claims the author makes directly.

  • The same flow-balance plus geometric correction may apply to other self-exciting product loops (referral bonuses, invite chains, collaborative play) whenever logs separate discovery-origin events from socially induced ones.
  • If platforms store generation depth or multi-hop attribution, one could test whether a generation-stratified rate estimator shrinks residual bias when homogeneity fails.
  • Heterogeneous or near-critical cascades (spectral radius close to one) are the regime where the method’s advantage over first-order adjustments should grow most, matching the paper’s stronger-propagation simulations.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, simulated authors' rebuttal, and a circularity audit.

Referee Report

3 major / 0 minor

Summary. The paper studies A/B evaluation of sharing features under multi-round social interference (the “flywheel”). It models share-induced views via a multivariate Hawkes process, derives a sender–receiver flow-balance identity, and interprets share-induced engagement as geometric amplification with effective downstream rate q_eff. This yields a log-based plug-in estimator [GTE = bY^d_T/(1−bq_T) − bY^d_C/(1−bq_C) for the global treatment effect on share-induced engagement, using attribution logs under Bernoulli randomization. Under Assumptions A1–A2 the estimator is consistent (Theorem 1); under A/A tests a finite-population CLT and plug-in SE give asymptotic Type I control (Theorem 2). Simulations (n=50k, multiple graphs/parameters) show large bias reduction versus DM, DM-FO, and GCR; a Poisson-heuristic extension targets reactivation; a real-platform A/A is roughly uniform and one A/B finds a significant effect where baselines do not.

Significance. If the method is reliable beyond the homogeneous-q regime, it fills a clear gap: interference from multi-round sharing is practically important and under-studied relative to marketplace and neighborhood interference. Strengths include a transparent flow-balance derivation that does not require fitting Hawkes kernels, closed-form propagation adjustment from standard attribution logs, machine-checkable-style proofs in Appendix A, public simulation code, extensive robustness sweeps (topology, density, parameter laws, spectral radius), and real deployment with A/A pipeline validation. The geometric representation is simple enough for production experimentation stacks. The main scientific value is a practical GTE estimator tailored to share cascades rather than generic exposure mappings.

major comments (3)
  1. Theorem 1 establishes consistency of [GTE only under Assumption A2 (homogeneous downstream rate q^a_ik ≡ q_a for all (i,k) within each regime). The estimator plugs a single bq_a = bY^s_a/cW^s_a into the geometric factor 1/(1−bq_a). Section 4.3 correctly flags A2 as stronger than necessary and appeals to aggregation plus simulations that already violate A2 via sender-side δ_i. That is supportive but not a proof that the common-q representation of GTE remains approximately unbiased under systematic receiver/content/treatment heterogeneity (e.g., treated senders preferentially share high-virality content, or high-degree receivers have higher q). For the central practical claim, please either (i) prove consistency/approximate unbiasedness under a weaker average-q or L2-heterogeneity condition, or (ii) add a targeted bias analysis/simulation design where q_ik varies with treatment, content, a
  2. Theorem 2 and the SE in Eq. (11) are justified only under A/A tests (identical kernels across arms). Table 1 reports near-nominal 95% coverage in A/B simulations, and Table 2 reports A/B p-values (e.g., 0.005 for the proposed method) that drive the launch decision narrative in §8. The manuscript does not state conditions under which bse remains valid when treatment changes dd, ds and thus the joint law of (Y^d, Y^s, W^s). Please either extend the CLT/SE theory to local alternatives / A/B under A1–A2, or clearly label Table 2 A/B inference as heuristic, report bootstrap or design-based alternatives, and temper the claim that the feature is “statistically significant” solely on the A/A-derived SE.
  3. Section 6’s reactivation estimator [GTE_ra = exp(−bλ_C)−exp(−bλ_T) is presented as a Poisson approximation with no consistency theorem, and Figure 2 shows remaining bias (though lower MSE than EW/HEW). The abstract and contributions list this as a framework extension on equal footing with the IS estimator. Please either supply conditions under which the Poisson plug-in is consistent for GTEra, or reframe §6 as an exploratory heuristic and avoid implying the same guarantees as Theorem 1 for the reactivation metric.

Circularity Check

0 steps flagged

No circularity: plug-in GTE follows from a proven flow-balance identity under stated assumptions; simulation ground truth and platform A/A–A/B checks are independent of the estimator.

full rationale

The derivation chain is self-contained and non-circular. The target GTE is defined as IST − ISC from the Hawkes model (global regimes), not from the experimental plug-in. Proposition 2 and the flow-balance identity (3)–(5) are obtained from the cluster representation and conditional means of offspring, then rearranged under A2 into the geometric form IS = μd/(1−q). The estimator [GTE = bYd_T/(1−bq_T) − bYd_C/(1−bq_C) is a standard plug-in of sample means for those population quantities; consistency (Theorem 1) is proved under A1–A2 rather than assumed. Simulation ground truth is computed from separate full-treatment and full-control Hawkes runs, not by recycling the estimator. A/A CLT (Theorem 2) and real-platform A/A uniformity / A/B significance are external checks. The definitional existence of q_eff that makes the identity hold is ordinary reparameterization, not a self-definitional prediction. No load-bearing self-citation uniqueness claim, fitted-input-as-prediction, or renamed known result forces the central claim. Weakness of A2 is an assumption/correctness issue, not circularity.

Axiom & Free-Parameter Ledger

0 free parameters · 6 axioms · 1 invented entities

The central claim rests on a standard point-process model of sharing, a combinatorial flow-balance identity, two technical assumptions that keep cascades subcritical and (for consistency) homogeneous, and a Bernoulli experimental design. No free parameters are fitted to produce the main GTE estimator; the method is intentionally plug-in and log-based. The reactivation extension adds a Poisson approximation that is heuristic rather than derived.

axioms (6)
  • domain assumption Share and view events follow a multivariate Hawkes process with discovery intensity μ_ik and marked kernels φ^d, φ^s constrained by the social graph (Section 3.1).
    Standard modeling choice for social diffusion; used to justify the cluster representation and conditional means of offspring counts.
  • standard math Flow-balance identity: every share-induced view has exactly one sender and one receiver, so sender-side offspring counts equal receiver-side share-induced counts (Eq. 3).
    Combinatorial identity from the branching representation; holds pathwise under the model.
  • domain assumption Assumption A1: uniform bounds on discovery intensity and per-user share traffic, spectral-radius-style subcriticality (max row sums ≤ c̄ < 1), and non-degenerate average discovery-driven sharing (Section 3.2).
    Ensures IS is finite and sample averages concentrate; standard for stable Hawkes cascades.
  • ad hoc to paper Assumption A2: homogeneous downstream sharing rate q^a_ik ≡ q_a for all (i,k) within each regime a ∈ {T,C} (Section 4.3).
    Stronger than needed for the identity; required for the consistency proof of the plug-in q estimator. Authors note aggregation may dampen violations but do not prove robustness.
  • domain assumption Bernoulli exposure and treatment assignment with fixed π, p; intervention changes only sender-side propagation strengths (Section 4.1).
    Matches common layered experimentation systems; defines the observed experimental law.
  • ad hoc to paper For reactivation: per-edge success probabilities small enough that W^s_i is approximately Poisson (Section 6).
    Heuristic used only for the RA extension; not needed for the main IS estimator.
invented entities (1)
  • Effective downstream sharing rate q_eff and geometric amplification factor 1/(1−q_eff) no independent evidence
    purpose: Rewrites total share-induced engagement as discovery-driven shares times a closed-form multi-round multiplier, motivating the plug-in estimator.
    Definitional construct derived from the flow-balance identity under homogeneity; not an independent physical entity, but the key modeling abstraction the estimator rests on.

pith-pipeline@v1.1.0-grok45 · 27020 in / 3323 out tokens · 48535 ms · 2026-07-14T09:00:09.246630+00:00 · methodology

0 comments
read the original abstract

Sustainable user growth in online platforms depends not only on acquiring new users but also on reactivating and engaging existing ones through social sharing features. A well-designed sharing feature can trigger a self-reinforcing ``flywheel effect'': reactivated users become potential sharers whose engagement propagates through the network over multiple rounds, amplifying total engagement. Measuring the causal impact of such sharing features is challenging, as their effects unfold through complex social networks and temporal cascades, violating the no-interference assumption underlying classical A/B testing. We develop a framework for experiments on sharing features that accounts for interference caused by the flywheel effect and targets a global treatment effect on share-induced engagement. Our estimator is motivated by a flow-balance identity and interprets share-induced engagement as a geometric amplification process, yielding a closed-form propagation adjustment that accounts for multi-round diffusion using commonly available attribution logs. Under mild conditions, we establish consistency of the proposed estimator and develop a valid A/A testing procedure for pipeline validation. Simulation studies show that our method substantially reduces bias relative to the difference-in-means estimator and first-order adjustments, while the proposed A/A test maintains nominal Type I error. We also extend the framework to a user-level reactivation metric via a Poisson approximation. Finally, we demonstrate the approach on a real-world large-scale online platform and discuss empirical implications for evaluating sharing feature designs.

Figures

Figures reproduced from arXiv: 2607.10820 by Nian Si, Weitao Cheng, Yilin Li, Yong Wang.

Figure 1
Figure 1. Figure 1: Performance of different estimators for the impact of sharing under varying disturbance [PITH_FULL_IMAGE:figures/full_fig_p015_1.png] view at source ↗
Figure 2
Figure 2. Figure 2: Performance of different estimators for the reactivation rate under varying disturbance [PITH_FULL_IMAGE:figures/full_fig_p015_2.png] view at source ↗
Figure 3
Figure 3. Figure 3: Distribution of p-values for 200 A/A experiments. We further demonstrate the practical value of our approach through a real A/B experiment. In this experiment, the treatment group was exposed to a redesigned interface intended to encourage sharing and downstream engagement with shared content, while the control group retained the original design. We focus on the impact-of-sharing metrics (Eq. 2). By constr… view at source ↗
Figure 4
Figure 4. Figure 4: Performance of different estimators for the impact of sharing under varying exposure [PITH_FULL_IMAGE:figures/full_fig_p027_4.png] view at source ↗

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.

Reference graph

Works this paper leans on

62 extracted references · 6 linked inside Pith

  1. [1]

    The impact of sharing mechanism design on content sharing in online social networks

    Irina Heimbach and Oliver Hinz. The impact of sharing mechanism design on content sharing in online social networks. Information Systems Research , 29(3):592–611, 2018

    2018

  2. [2]

    The Value Flywheel Effect: Power the Future and Accelerate Your Organi- zation to the Modern Cloud

    David Anderson. The Value Flywheel Effect: Power the Future and Accelerate Your Organi- zation to the Modern Cloud . IT Revolution, 2022

    2022

  3. [3]

    Erdogdu, Hera Y

    Qingyuan Zhao, Murat A. Erdogdu, Hera Y. He, Anand Rajaraman, and Jure Leskovec. SEISMIC: A self-exciting point process model for predicting tweet popularity. In Proceedings of the 21st ACM SIGKDD International Conference on Knowledge Discovery and Data Mining (KDD), pages 1513–1522, 2015

    2015

  4. [4]

    A tutorial on hawkes processes for events in social media

    Marian-Andrei Rizoiu, Young Lee, Swapnil Mishra, and Lexing Xie. A tutorial on hawkes processes for events in social media. arXiv preprint arXiv:1708.06401 , 2017. 17

    Pith/arXiv arXiv 2017

  5. [5]

    Causal Inference in Statistics, Social, and Biomedical Sciences

    Guido W Imbens and Donald B Rubin. Causal Inference in Statistics, Social, and Biomedical Sciences. Cambridge University Press, 2015

    2015

  6. [6]

    Multivariate hawkes processes: an application to financial data

    Paul Embrechts, Thomas Liniger, and Lu Lin. Multivariate hawkes processes: an application to financial data. Journal of Applied Probability , 48(A):367–378, 2011

    2011

  7. [7]

    Social influence in social advertis- ing: evidence from field experiments

    Eytan Bakshy, Dean Eckles, Rong Yan, and Itamar Rosenn. Social influence in social advertis- ing: evidence from field experiments. In Proceedings of the 13th ACM conference on electronic commerce, pages 146–161, 2012

    2012

  8. [8]

    Unveiling environmental sensitivity of individual gains in influence maximization

    Xinyan Su, Zhiheng Zhang, and Jiyan Qiu. Unveiling environmental sensitivity of individual gains in influence maximization. arXiv preprint arXiv:2301.12226 , 2023

    Pith/arXiv arXiv 2023

  9. [9]

    Why marketplace experimentation is harder than it seems: The role of test-control interference

    Thomas Blake and Dominic Coey. Why marketplace experimentation is harder than it seems: The role of test-control interference. In Proceedings of the fifteenth ACM conference on Eco- nomics and computation , pages 567–582, 2014

    2014

  10. [10]

    Reducing interference bias in online marketplace pricing experiments

    David Holtz, Ruben Lobel, Inessa Liskovich, and Sinan Aral. Reducing interference bias in online marketplace pricing experiments. arXiv preprint arXiv:2004.12489 , 2020

    Pith/arXiv arXiv 2004

  11. [11]

    Search frictions and the design of online marketplaces

    Andrey Fradkin. Search frictions and the design of online marketplaces. In The Third Con- ference on Auctions, Market Mechanisms and Their Applications , 2015

    2015

  12. [12]

    Experimental design in two-sided platforms: An analysis of bias

    Ramesh Johari, Hannah Li, Inessa Liskovich, and Gabriel Y Weintraub. Experimental design in two-sided platforms: An analysis of bias. Management Science , 68(10):7069–7089, 2022

    2022

  13. [13]

    Interference, bias, and variance in two-sided marketplace experimentation: Guidance for platforms

    Hannah Li, Geng Zhao, Ramesh Johari, and Gabriel Y Weintraub. Interference, bias, and variance in two-sided marketplace experimentation: Guidance for platforms. In Proceedings of the ACM Web Conference 2022 , pages 182–192, 2022

    2022

  14. [14]

    Price experimentation and interference in online platforms

    Wassim Dhaouadi, Ramesh Johari, and Gabriel Y Weintraub. Price experimentation and interference in online platforms. arXiv preprint arXiv:2310.17165 , 2023

    arXiv 2023

  15. [15]

    Experimental design in marketplaces

    Patrick Bajari, Brian Burdick, Guido W Imbens, Lorenzo Masoero, James McQueen, Thomas S Richardson, and Ido M Rosen. Experimental design in marketplaces. Statistical Science , 38 (3):458–476, 2023

    2023

  16. [16]

    Multiple randomization designs: Estimation and inference with interference

    Lorenzo Masoero, Suhas Vijaykumar, Thomas S Richardson, James McQueen, Ido Rosen, Brian Burdick, Pat Bajari, and Guido Imbens. Multiple randomization designs: Estimation and inference with interference. Journal of the Royal Statistical Society Series B: Statistical Methodology, page qkaf073, 2026

    2026

  17. [17]

    Reducing marketplace interference bias via shadow prices

    Ido Bright, Arthur Delarue, and Ilan Lobel. Reducing marketplace interference bias via shadow prices. Management Science , 71(8):7094–7112, 2025

    2025

  18. [18]

    Experimental design for matching

    Chonghuan Wang. Experimental design for matching. arXiv preprint arXiv:2601.21036 , 2026

    arXiv 2026

  19. [19]

    A/B testing of auctions

    Shuchi Chawla, Jason Hartline, and Denis Nekipelov. A/B testing of auctions. In Proceedings of the 2016 ACM Conference on Economics and Computation , pages 19–20, 2016

    2016

  20. [20]

    Randomization and the pernicious effects of limited budgets on auction experiments

    Guillaume W Basse, Hossein Azari Soufiani, and Diane Lambert. Randomization and the pernicious effects of limited budgets on auction experiments. In Artificial Intelligence and Statistics, pages 1412–1420. PMLR, 2016. 18

    2016

  21. [21]

    Statistical inference and a/b testing for first-price pacing equilibria

    Luofeng Liao and Christian Kroer. Statistical inference and a/b testing for first-price pacing equilibria. In International Conference on Machine Learning , pages 20868–20905. PMLR, 2023

    2023

  22. [22]

    Interference among first-price pacing equilibria: A bias and variance analysis

    Luofeng Liao, Christian Kroer, Sergei Leonenkov, Okke Schrijvers, Liang Shi, Nicolas Stier- Moses, and Congshan Zhang. Interference among first-price pacing equilibria: A bias and variance analysis. arXiv preprint arXiv:2402.07322 , 2024

    Pith/arXiv arXiv 2024

  23. [23]

    Tackling interference induced by data training loops in a/b tests: A weighted training approach

    Nian Si. Tackling interference induced by data training loops in a/b tests: A weighted training approach. arXiv preprint arXiv:2310.17496 , 2023

    Pith/arXiv arXiv 2023

  24. [24]

    Estimating treatment effects under recommender interference: A structured neural networks approach

    Ruohan Zhan, Shichao Han, Yuchen Hu, and Zhenling Jiang. Estimating treatment effects under recommender interference: A structured neural networks approach. arXiv preprint arXiv:2406.14380, 2024

    arXiv 2024

  25. [25]

    Debiasing seller-side experiments under multinomial logit models in two-sided platforms

    Chenyu Zhang, Yuhang Wu, Zeyu Zheng, and Nian Si. Debiasing seller-side experiments under multinomial logit models in two-sided platforms. Available at SSRN 5261661 , 2025

    2025

  26. [26]

    Causal inference with dyadic data in ran- domized experiments

    Yilin Li, Lu Deng, Yong Wang, and Wang Miao. Causal inference with dyadic data in ran- domized experiments. arXiv preprint arXiv:2505.20780 , 2025

    Pith/arXiv arXiv 2025

  27. [27]

    Detecting interference using dyadic data in online con- trolled experiments

    Yilin Li, Lu Deng, and Yong Wang. Detecting interference using dyadic data in online con- trolled experiments. In Proceedings of the 31st ACM SIGKDD Conference on Knowledge Discovery and Data Mining , pages 1623–1634, 2025

    2025

  28. [28]

    Experimenting in equilibrium

    Stefan Wager and Kuang Xu. Experimenting in equilibrium. Management Science , 67(11): 6694–6715, 2021

    2021

  29. [29]

    Treatment effects in market equilibrium

    Evan Munro, Xu Kuang, and Stefan Wager. Treatment effects in market equilibrium. American Economic Review, 115(10):3273–3321, 2025

    2025

  30. [30]

    Toward causal inference with interference

    Michael G Hudgens and M Elizabeth Halloran. Toward causal inference with interference. Journal of the American Statistical Association , 103(482):832–842, 2008

    2008

  31. [31]

    Network a/b testing: From sampling to estimation

    Huan Gui, Ya Xu, Anmol Bhasin, and Jiawei Han. Network a/b testing: From sampling to estimation. In Proceedings of the 24th International Conference on World Wide Web , pages 399–409, 2015

    2015

  32. [32]

    Randomization tests of causal effects under interference

    Guillaume W Basse, A vi Feller, and Panos Toulis. Randomization tests of causal effects under interference. Biometrika, 106(2):487–494, 2019

    2019

  33. [33]

    Random graph asymptotics for treatment effect estimation under network interference

    Shuangning Li and Stefan Wager. Random graph asymptotics for treatment effect estimation under network interference. The Annals of Statistics , 50(4):2334–2358, 2022

    2022

  34. [34]

    Treatment effect accounting for network changes

    Margherita Comola and Silvia Prina. Treatment effect accounting for network changes. Review of Economics and Statistics , 103(3):597–604, 2021

    2021

  35. [35]

    Causal message-passing for experiments with unknown and general network interference

    Sadegh Shirani and Mohsen Bayati. Causal message-passing for experiments with unknown and general network interference. Proceedings of the National Academy of Sciences , 121(40): e2322232121, 2024

    2024

  36. [36]

    Estimating average causal effects under general interference, with application to a social network experiment

    Peter M Aronow and Cyrus Samii. Estimating average causal effects under general interference, with application to a social network experiment. The Annals of Applied Statistics , 11(4):1912– 1947, 2017. 19

    1912

  37. [37]

    Correlated cluster-based randomized exper- iments: Robust variance minimization

    Ozan Candogan, Chen Chen, and Rad Niazadeh. Correlated cluster-based randomized exper- iments: Robust variance minimization. Management Science , 70(6):4069–4086, 2024

    2024

  38. [38]

    Graph cluster random- ization: Network exposure to multiple universes

    Johan Ugander, Brian Karrer, Lars Backstrom, and Jon Kleinberg. Graph cluster random- ization: Network exposure to multiple universes. In Proceedings of the 19th ACM SIGKDD International Conference on Knowledge Discovery and Data Mining , pages 329–337, 2013

    2013

  39. [39]

    Randomized graph cluster randomization

    Johan Ugander and Hao Yin. Randomized graph cluster randomization. Journal of Causal Inference, 11(1):20220014, 2023

    2023

  40. [40]

    Experimental design in one-sided matching platforms

    Chenran Weng, Xiao Lei, and Nian Si. Experimental design in one-sided matching platforms. Available at SSRN 4890353 , 2024

    2024

  41. [41]

    Estimation of causal peer influence effects

    Panos Toulis and Edward Kao. Estimation of causal peer influence effects. In International conference on machine learning , pages 1489–1497. PMLR, 2013

    2013

  42. [42]

    Optimal design of experiments in the presence of interference

    Sarah Baird, J Aislinn Bohren, Craig McIntosh, and Berk Özler. Optimal design of experiments in the presence of interference. Review of Economics and Statistics , 100(5):844–860, 2018

    2018

  43. [43]

    Identification of peer effects through social networks

    Yann Bramoullé, Habiba Djebbari, and Bernard Fortin. Identification of peer effects through social networks. Journal of econometrics , 150(1):41–55, 2009

    2009

  44. [44]

    Design and analysis of bipartite experiments under a linear exposure-response model

    Christopher Harshaw, Fredrik Sävje, David Eisenstat, Vahab Mirrokni, and Jean Pouget- Abadie. Design and analysis of bipartite experiments under a linear exposure-response model. Electronic Journal of Statistics , 17(1):464–518, 2023

    2023

  45. [45]

    Estimating the total treatment effect in randomized experiments with unknown network structure

    Christina Lee Yu, Edoardo M Airoldi, Christian Borgs, and Jennifer T Chayes. Estimating the total treatment effect in randomized experiments with unknown network structure. Proceedings of the National Academy of Sciences , 119(44):e2208975119, 2022

    2022

  46. [46]

    Estimating causal effects under network interference with bayesian generalized propensity scores

    Laura Forastiere, Fabrizia Mealli, Albert Wu, and Edoardo M Airoldi. Estimating causal effects under network interference with bayesian generalized propensity scores. Journal of Machine Learning Research, 23(289):1–61, 2022

    2022

  47. [47]

    Causal inference under approximate neighborhood interference

    Michael P Leung. Causal inference under approximate neighborhood interference. Economet- rica, 90(1):267–293, 2022

    2022

  48. [48]

    Randomization tests for peer effects in group formation experiments

    Guillaume Basse, Peng Ding, A vi Feller, and Panos Toulis. Randomization tests for peer effects in group formation experiments. Econometrica, 92(2):567–590, 2024

    2024

  49. [49]

    Expect- ing to be HIP: Hawkes intensity processes for social media popularity

    Marian-Andrei Rizoiu, Swapnil Mishra, Quang Kong, Mark Carman, and Lexing Xie. Expect- ing to be HIP: Hawkes intensity processes for social media popularity. In Proceedings of the International Conference on World Wide Web (WWW) , 2017

    2017

  50. [50]

    Shaping social activity by incentivizing users

    Mehrdad Farajtabar, Nan Du, Manuel Gomez-Rodriguez, Isabel Valera, Hongyuan Zha, and Le Song. Shaping social activity by incentivizing users. In Advances in Neural Information Processing Systems (NIPS) , 2014

    2014

  51. [51]

    An introduction to the theory of point processes: volume I: elementary theory and methods

    Daryl J Daley and David Vere-Jones. An introduction to the theory of point processes: volume I: elementary theory and methods . Springer, 2003

    2003

  52. [52]

    Multivariate Hawkes Processes

    Thomas Liniger. Multivariate Hawkes Processes . PhD thesis, ETH Zurich, 2009. 20

    2009

  53. [53]

    Uncovering causality from multivariate hawkes integrated cumulants

    Massil Achab, Emmanuel Bacry, Stéphane Gaïffas, Iacopo Mastromatteo, and Jean-François Muzy. Uncovering causality from multivariate hawkes integrated cumulants. Journal of Ma- chine Learning Research, 18(192):1–28, 2018

    2018

  54. [54]

    A cluster process representation of a self-exciting process

    Alan G Hawkes and David Oakes. A cluster process representation of a self-exciting process. Journal of applied probability , 11(3):493–503, 1974

    1974

  55. [55]

    From Louvain to Leiden: guaran- teeing well-connected communities

    Vincent A Traag, Ludo Waltman, and Nees Jan Van Eck. From Louvain to Leiden: guaran- teeing well-connected communities. Scientific Reports, 9(1):1–12, 2019

    2019

  56. [56]

    Statistical mechanics of complex networks

    Réka Albert and Albert-László Barabási. Statistical mechanics of complex networks. Reviews of Modern Physics , 74(1):47, 2002

    2002

  57. [57]

    Detecting interference in online controlled experiments with increasing allocation

    Kevin Han, Shuangning Li, Jialiang Mao, and Han Wu. Detecting interference in online controlled experiments with increasing allocation. In Proceedings of the 29th ACM SIGKDD Conference on Knowledge Discovery and Data Mining , pages 661–672, 2023

    2023

  58. [58]

    Branching processes

    Krishna B Athreya and Peter E Ney. Branching processes. Springer Science & Business Media, 2012

    2012

  59. [59]

    A first course in causal inference

    Peng Ding. A first course in causal inference . Chapman and Hall/CRC, 2024

    2024

  60. [60]

    General forms of finite population central limit theorems with applications to causal inference

    Xinran Li and Peng Ding. General forms of finite population central limit theorems with applications to causal inference. Journal of the American Statistical Association , 112(520): 1759–1769, 2017

    2017

  61. [61]

    Watts and Steven H

    Duncan J. Watts and Steven H. Strogatz. Collective dynamics of ‘small-world’ networks. Nature, 393(6684):440–442, 1998

    1998

  62. [62]

    Holland, Kathryn Blackmond Laskey, and Samuel Leinhardt

    Paul W. Holland, Kathryn Blackmond Laskey, and Samuel Leinhardt. Stochastic blockmodels: First steps. Social Networks , 5(2):109–137, 1983. 21 A Proofs Proof of Proposition 1. Under Assumption A1(2), ∥Q(k)∥∞ ≤ ¯c < 1 for every k. Hence a single discovery-driven view event of content k triggers, in expectation, at most X ℓ≥1 ∥Q(k)∥ℓ ∞ ≤ ¯c 1 − ¯c share-ind...

    1983