Challenges
Datasets
Workspaces
Papers
Discussions
Leaderboard
Log inSign up
Challenges
Datasets
Workspaces
Papers
Discussions
Leaderboard
Blog
Job Board
Q3AS

© 2026 Aqora Quantum S.A.S.

TermsPrivacyLegal Notice

arXiv:1203.1114 · Journal of Physics A Mathematical and Theoretical

Two quantum Simpson’s paradoxes

Matteo G A Paris

Published Mar 15, 2012• 11 citations
arXivHTMLPDF

Abstract

The so-called Simpson's 'paradox', or Yule–Simpson (YS) effect, occurs in classical statistics when the correlations that are present among different sets of samples are reversed if the sets are combined together, thus ignoring one or more lurking variables. Here we illustrate the occurrence of two analog effects in quantum measurements. The first, which we term the quantum–classical YS effect, may occur with quantum limited measurements and with lurking variables coming from the mixing of states, whereas the second, here referred to as the quantum–quantum YS effect, may take place when coherent superpositions of quantum states are allowed. By analyzing quantum measurements on low-dimensional systems (qubits and qutrits), we show that the two effects may occur independently and that the quantum–quantum YS effect is more likely to occur than the corresponding quantum–classical one. We also found that there exist classes of superposition states for which the quantum–classical YS effect cannot occur for any measurement and, at the same time, the quantum–quantum YS effect takes place in a consistent fraction of the possible measurement settings. The occurrence of the effect in the presence of partial coherence is discussed as well as its possible implications for quantum hypothesis testing.

DOI
10.1088/1751-8113/45/13/132001

Related papers

Experimental investigation of quantum Simpson's paradox

Yulong Li, Jian‐Shun Tang, Yi‐Tao Wang +9 more

The well-known Simpson's paradox, or Yule-Simpson (YS) effect, is often encountered in social-science and medical-science statistics. It occurs when the correlations present in different groups are reversed if the groups are combined. Simpson's paradox also exists in quantum measurements. In this Brief Report, we experimentally realized two analogous effects: the quantum-classical YS effect and the quantum-quantum YS effect in the quantum-dot system. We also compared the probability of obtaining those two effects under identical quantum measurements and found that the quantum-quantum YS effect is more likely to occur than the quantum-classical YS effect.

Physical Review A•Jul 24, 2013•9 citations

The Simpson’s paradox in quantum mechanics

Alessandro Selvitella

In probability and statistics, the Simpson’s paradox is a paradox in which a trend that appears in different groups of data disappears when these groups are combined, while the reverse trend appears for the aggregate data. In this paper, we give some results about the occurrence of the Simpson’s paradox in quantum mechanics. In particular, we prove that the Simpson’s paradox occurs for solutions of the quantum harmonic oscillator both in the stationary case and in the non-stationary case. In the non-stationary case, the Simpson’s paradox is persistent: if it occurs at any time t=t∼, then it occurs at any time t≠t∼. Moreover, we prove that the Simpson’s paradox is not an isolated phenomenon, namely, that, close to initial data for which it occurs, there are lots of initial data (a open neighborhood), for which it still occurs. Differently from the case of the quantum harmonic oscillator, we also prove that the paradox appears (asymptotically) in the context of the nonlinear Schrödinger equation but at intermittent times.

Journal of Mathematical Physics•Mar 1, 2017•6 citations

Boson correlations are spurious for classical states

Daniel E. Salazar, Fabrice P. Laussy

We show that boson correlations from quantum states with a Glauber-Sudarshan representation of their density matrix which provides a well-behaved probability distribution -- including coherent states, thermal states, and all states that can be deemed classical -- are a manifestation of the Simpson paradox: they are spurious correlations from statistical (ensemble) averages over uncorrelated measurements made in varying geometries, due to a process of symmetry-breaking as a confounding factor. Bosonic correlations encoded by the wavefunction appear to be formed in the geometry assumed, which however is not that of the statistical ensemble but varies from realization to realization. This calls to distinguish between quantum and statistical averages and sheds new understandings on the fundamental problems of nonclassicality and quantum advantage.

quant-pharXiv:2604.16283•Physical Review Letters•Apr 20, 2026•1 citation

Quantum Simpsons Paradox and High Order Bell-Tsirelson Inequalities

Yaoyun Shi

The well-known Simpson's Paradox, or Yule-Simpson Effect, in statistics is often illustrated by the following thought experiment: A drug may be found in a trial to increase the survival rate for both men and women, but decrease the rate for all the subjects as a whole. This paradoxical reversal effect has been found in numerous datasets across many disciplines, and is now included in most introductory statistics textbooks. In the language of the drug trial, the effect is impossible, however, if both treatment groups' survival rates are higher than both control groups'. Here we show that for quantum probabilities, such a reversal remains possible. In particular, a "quantum drug", so to speak, could be life-saving for both men and women yet deadly for the whole population. We further identify a simple inequality on conditional probabilities that must hold classically but is violated by our quantum scenarios, and completely characterize the maximum quantum violation. As polynomial inequalities on entries of the density operator, our inequalities are of degree 6.

arXiv:1203.2675•Mar 12, 2012•3 citations

Simpson’s paradox in cross-platform quantum deviceanalysis: A methodological case study using Josephsonjunction decoherence data

Mahgoub A. Salih

Abstract Cross-platform comparison of quantum devices requires careful statistical methodologyto avoid spurious conclusions. We present a methodological case study demonstrating howSimpson’s paradox can arise in meta-analyses of superconducting circuit data. We introducetwo physically motivated scaling parameters—a dimensionless normalized temperature T∗ =kBT/hf and a resistance-like decoherence parameter A = (hf/eIc)(ln Q/2π)—and applythem to a secondary analysis of published data from Transmon, SNS, and SINIS devices(N = 10 data points from 3 studies).Across the pooled dataset, we observe a strong correlation (r = 0.910) between A and T∗.However, we demonstrate that this correlation is a statistical artifact arising from completeconfounding between temperature and device type: the three platforms operate in nonoverlappingtemperature regimes, and within-group correlations are consistently negative(though not statistically significant given the small sample sizes of n = 3–4 per group).Mixed-effects modeling indicates that platform identity accounts for the dominant share ofvariance in A (ICC ≈ 0.98), though we caution that variance component estimates from onlyk = 3 groups carry substantial uncertainty.Our analysis serves as a methodological warning for cross-platform data aggregation inquantum physics, illustrating how pooled correlations can reverse direction when data aredisaggregated—a hallmark of Simpson’s paradox. The proposed scaling framework may proveuseful for future benchmarking, but we emphasize that the current data are insufficient totest any hypothesis of universal thermodynamic scaling. We outline experimental conditionsnecessary for such tests, including overlapping temperature ranges and controlled singledevicesweeps.This work is intended as a methodological case study rather than a test of aphysical scaling law

•Feb 4, 2026•0 citations

Conditions for logical contextuality and nonlocality

Leonardo Santos, Bárbara Amaral

Contextuality and nonlocality are nonclassical properties exhibited by quantum statistics whose implications profoundly impact both the foundations and applications of quantum theory. In this paper we provide some insights into logical contextuality and inequality-free proofs. The former can be understood as the possibility version of contextuality, while the latter refers to proofs of quantum contextuality and nonlocality that are not based on violations of some noncontextuality (or Bell) inequality. By ``possibilistic'' we mean a description in terms of possibilities for the outcomes, which are Boolean variables assuming value one when the corresponding probability is strictly larger than zero and zero otherwise. The present work aims to build a bridge between these two concepts from what we call possibilistic paradoxes, which are sets of possibilistic conditions whose occurrence implies contextuality and nonlocality. As the main result, we demonstrate the existence of possibilistic paradoxes whose occurrence is a necessary and sufficient condition for logical contextuality in a very important class of scenarios. Finally, we discuss some interesting consequences arising from the completeness of these possibilistic paradoxes.

arXiv:2011.04111•Physical Review A•Aug 4, 2021•10 citations
The so-called Simpson's 'paradox', or Yule–Simpson (YS) effect, occurs in classical statistics when the correlations that are present among different sets of samples are reversed if the sets are combined together, thus ignoring one or more lurking variables. Here we illustrate the occurrence of two analog effects in quantum measurements. The first, which we term the quantum–classical YS effect, may occur with quantum limited measurements and with lurking variables coming from the mixing of states, whereas the second, here referred to as the quantum–quantum YS effect, may take place when coherent superpositions of quantum states are allowed. By analyzing quantum measurements on low-dimensional systems (qubits and qutrits), we show that the two effects may occur independently and that the quantum–quantum YS effect is more likely to occur than the corresponding quantum–classical one. We also found that there exist classes of superposition states for which the quantum–classical YS effect cannot occur for any measurement and, at the same time, the quantum–quantum YS effect takes place in a consistent fraction of the possible measurement settings. The occurrence of the effect in the presence of partial coherence is discussed as well as its possible implications for quantum hypothesis testing.
The well-known Simpson's paradox, or Yule-Simpson (YS) effect, is often encountered in social-science and medical-science statistics. It occurs when the correlations present in different groups are reversed if the groups are combined. Simpson's paradox also exists in quantum measurements. In this Brief Report, we experimentally realized two analogous effects: the quantum-classical YS effect and the quantum-quantum YS effect in the quantum-dot system. We also compared the probability of obtaining those two effects under identical quantum measurements and found that the quantum-quantum YS effect is more likely to occur than the quantum-classical YS effect.
In probability and statistics, the Simpson’s paradox is a paradox in which a trend that appears in different groups of data disappears when these groups are combined, while the reverse trend appears for the aggregate data. In this paper, we give some results about the occurrence of the Simpson’s paradox in quantum mechanics. In particular, we prove that the Simpson’s paradox occurs for solutions of the quantum harmonic oscillator both in the stationary case and in the non-stationary case. In the non-stationary case, the Simpson’s paradox is persistent: if it occurs at any time t=t∼, then it occurs at any time t≠t∼. Moreover, we prove that the Simpson’s paradox is not an isolated phenomenon, namely, that, close to initial data for which it occurs, there are lots of initial data (a open neighborhood), for which it still occurs. Differently from the case of the quantum harmonic oscillator, we also prove that the paradox appears (asymptotically) in the context of the nonlinear Schrödinger equation but at intermittent times.
We show that boson correlations from quantum states with a Glauber-Sudarshan representation of their density matrix which provides a well-behaved probability distribution -- including coherent states, thermal states, and all states that can be deemed classical -- are a manifestation of the Simpson paradox: they are spurious correlations from statistical (ensemble) averages over uncorrelated measurements made in varying geometries, due to a process of symmetry-breaking as a confounding factor. Bosonic correlations encoded by the wavefunction appear to be formed in the geometry assumed, which however is not that of the statistical ensemble but varies from realization to realization. This calls to distinguish between quantum and statistical averages and sheds new understandings on the fundamental problems of nonclassicality and quantum advantage.
The well-known Simpson's Paradox, or Yule-Simpson Effect, in statistics is often illustrated by the following thought experiment: A drug may be found in a trial to increase the survival rate for both men and women, but decrease the rate for all the subjects as a whole. This paradoxical reversal effect has been found in numerous datasets across many disciplines, and is now included in most introductory statistics textbooks. In the language of the drug trial, the effect is impossible, however, if both treatment groups' survival rates are higher than both control groups'. Here we show that for quantum probabilities, such a reversal remains possible. In particular, a "quantum drug", so to speak, could be life-saving for both men and women yet deadly for the whole population. We further identify a simple inequality on conditional probabilities that must hold classically but is violated by our quantum scenarios, and completely characterize the maximum quantum violation. As polynomial inequalities on entries of the density operator, our inequalities are of degree 6.
Abstract Cross-platform comparison of quantum devices requires careful statistical methodologyto avoid spurious conclusions. We present a methodological case study demonstrating howSimpson’s paradox can arise in meta-analyses of superconducting circuit data. We introducetwo physically motivated scaling parameters—a dimensionless normalized temperature T∗ =kBT/hf and a resistance-like decoherence parameter A = (hf/eIc)(ln Q/2π)—and applythem to a secondary analysis of published data from Transmon, SNS, and SINIS devices(N = 10 data points from 3 studies).Across the pooled dataset, we observe a strong correlation (r = 0.910) between A and T∗.However, we demonstrate that this correlation is a statistical artifact arising from completeconfounding between temperature and device type: the three platforms operate in nonoverlappingtemperature regimes, and within-group correlations are consistently negative(though not statistically significant given the small sample sizes of n = 3–4 per group).Mixed-effects modeling indicates that platform identity accounts for the dominant share ofvariance in A (ICC ≈ 0.98), though we caution that variance component estimates from onlyk = 3 groups carry substantial uncertainty.Our analysis serves as a methodological warning for cross-platform data aggregation inquantum physics, illustrating how pooled correlations can reverse direction when data aredisaggregated—a hallmark of Simpson’s paradox. The proposed scaling framework may proveuseful for future benchmarking, but we emphasize that the current data are insufficient totest any hypothesis of universal thermodynamic scaling. We outline experimental conditionsnecessary for such tests, including overlapping temperature ranges and controlled singledevicesweeps.This work is intended as a methodological case study rather than a test of aphysical scaling law
Contextuality and nonlocality are nonclassical properties exhibited by quantum statistics whose implications profoundly impact both the foundations and applications of quantum theory. In this paper we provide some insights into logical contextuality and inequality-free proofs. The former can be understood as the possibility version of contextuality, while the latter refers to proofs of quantum contextuality and nonlocality that are not based on violations of some noncontextuality (or Bell) inequality. By ``possibilistic'' we mean a description in terms of possibilities for the outcomes, which are Boolean variables assuming value one when the corresponding probability is strictly larger than zero and zero otherwise. The present work aims to build a bridge between these two concepts from what we call possibilistic paradoxes, which are sets of possibilistic conditions whose occurrence implies contextuality and nonlocality. As the main result, we demonstrate the existence of possibilistic paradoxes whose occurrence is a necessary and sufficient condition for logical contextuality in a very important class of scenarios. Finally, we discuss some interesting consequences arising from the completeness of these possibilistic paradoxes.