scholarly journals Approximate Kernel-Based Conditional Independence Tests for Fast Non-Parametric Causal Discovery

2019 ◽  
Vol 7 (1) ◽  
Author(s):  
Eric V. Strobl ◽  
Kun Zhang ◽  
Shyam Visweswaran
Keyword(s):  
Sample Size ◽  
Conditional Independence ◽  
Causal Discovery ◽  
Large Sample Size ◽  
Conditional Correlation ◽  
Independence Test ◽  
P Values ◽  
Random Fourier Features ◽  
Non Parametric ◽  
Conditional Independence Test

AbstractConstraint-based causal discovery (CCD) algorithms require fast and accurate conditional independence (CI) testing. The Kernel Conditional Independence Test (KCIT) is currently one of the most popular CI tests in the non-parametric setting, but many investigators cannot use KCIT with large datasets because the test scales at least quadratically with sample size. We therefore devise two relaxations called the Randomized Conditional Independence Test (RCIT) and the Randomized conditional Correlation Test (RCoT) which both approximate KCIT by utilizing random Fourier features. In practice, both of the proposed tests scale linearly with sample size and return accurate p-values much faster than KCIT in the large sample size context. CCD algorithms run with RCIT or RCoT also return graphs at least as accurate as the same algorithms run with KCIT but with large reductions in run time.

scholarly journals Recursively Learning Causal Structures Using Regression-Based Conditional Independence Test

2019 ◽  
Vol 33 ◽  
pp. 3108-3115
Author(s):  
Hao Zhang ◽  
Shuigeng Zhou ◽  
Chuanxu Yan ◽  
Jihong Guan ◽  
Xin Wang
Keyword(s):  
Conditional Independence ◽  
Original Data ◽  
Equivalence Classes ◽  
Independence Test ◽  
Decomposition Approach ◽  
Decomposition Scheme ◽  
Causal Direction ◽  
Causality Inference ◽  
Non Gaussian ◽  
Conditional Independence Test

This paper addresses two important issues in causality inference. One is how to reduce redundant conditional independence (CI) tests, which heavily impact the efficiency and accuracy of existing constraint-based methods. Another is how to construct the true causal graph from a set of Markov equivalence classes returned by these methods.For the first issue, we design a recursive decomposition approach where the original data (a set of variables) is first decomposed into three small subsets, each of which is then recursively decomposed into three smaller subsets until none of subsets can be decomposed further. Consequently, redundant CI tests can be reduced by inferring causality from these subsets. Advantage of this decomposition scheme lies in two aspects: 1) it requires only low-order CI tests, and 2) it does not violate d-separation. Thus, the complete causality can be reconstructed by merging all the partial results of the subsets.For the second issue, we employ regression-based conditional independence test to check CIs in linear non-Gaussian additive noise cases, which can identify more causal directions by x−E(x|Z)⊥z (or y−E(y|Z)⊥z). Therefore, causal direction learning is no longer limited by the number of returned Vstructures and the consistent propagation.Extensive experiments show that the proposed method can not only substantially reduce redundant CI tests but also effectively distinguish the equivalence classes, thus is superior to the state of the art constraint-based methods in causality inference.

scholarly journals Bayesian interpretation of p values in clinical trials

2021 ◽  
pp. bmjebm-2020-111603
Author(s):  
John Ferguson
Keyword(s):  
Clinical Trial ◽  
Clinical Trials ◽  
Sample Size ◽  
Confidence Intervals ◽  
Statistical Significance ◽  
Large Sample Size ◽  
P Values ◽  
Clinical Trial Results ◽  
Sound Treatment ◽  
Counterintuitive Result

Commonly accepted statistical advice dictates that large-sample size and highly powered clinical trials generate more reliable evidence than trials with smaller sample sizes. This advice is generally sound: treatment effect estimates from larger trials tend to be more accurate, as witnessed by tighter confidence intervals in addition to reduced publication biases. Consider then two clinical trials testing the same treatment which result in the same p values, the trials being identical apart from differences in sample size. Assuming statistical significance, one might at first suspect that the larger trial offers stronger evidence that the treatment in question is truly effective. Yet, often precisely the opposite will be true. Here, we illustrate and explain this somewhat counterintuitive result and suggest some ramifications regarding interpretation and analysis of clinical trial results.

scholarly journals Feasibility of Race by Sex Intersectionality Research on Suicidality in the Adolescent Brain Cognitive Development (ABCD) Study

Children ◽  
2021 ◽  
Vol 8 (6) ◽  
pp. 437
Author(s):  
Shervin Assari ◽  
Shanika Boyce ◽  
Mohsen Bazargan
Keyword(s):  
Cognitive Development ◽  
Sample Size ◽  
Child Behavior Checklist ◽  
Reliability And Validity ◽  
Self Report ◽  
Large Sample Size ◽  
Cross Sectional ◽  
Cronbach Alpha ◽  
Rich Data ◽  
Suicidal Ideas

Intersectional research on childhood suicidality requires studies with a reliable and valid measure of suicidality, as well as a large sample size that shows some variability of suicidality across sex by race intersectional groups. Objectives: We aimed to investigate the feasibility of intersectionality research on childhood suicidality in the Adolescent Brain Cognitive Development (ABCD) study. We specifically explored the reliability and validity of the measure, sample size, and variability of suicidality across sex by race intersectional groups. Methods: We used cross-sectional data (wave 1) from the ABCD study, which sampled 9013 non-Hispanic white (NHW) or non-Hispanic black (NHB) children between the ages of 9 and 10 between years 2016 and 2018. Four intersectional groups were built based on race and sex: NHW males (n = 3554), NHW females (n = 3158), NHB males (n = 1164), and NHB females (n = 1137). Outcome measure was the count of suicidality symptoms, reflecting all positive history and symptoms of suicidal ideas, plans, and attempts. To validate our measure, we tested the correlation between our suicidality measure and depression and Child Behavior Checklist (CBCL) sub-scores. Cronbach alpha was calculated for reliability across each intersectional group. We also compared groups for suicidality. Results: We observed some suicidality history in observed 3.2% (n = 101) of NHW females, 4.9% (n = 175) of NHW males, 5.4% (n = 61) of NHB females, and 5.8% (n = 68) of NHB males. Our measure’s reliability was acceptable in all race by sex groups (Cronbach alpha higher than .70+ in all intersectional groups). Our measure was valid in all intersectional groups, documented by a positive correlation with depression and CBCL sub-scores. We could successfully model suicidality across sex by race groups, using multivariable models. Conclusion: Given the high sample size, reliability, and validity of the suicidality measure, variability of suicidality, it is feasible to investigate correlates of suicidality across race by sex intersections in the ABCD study. We also found evidence of higher suicidality in NHB than NHW children in the ABCD study. The ABCD rich data in domains of social context, self-report, schools, parenting, psychopathology, personality, and brain imaging provides a unique opportunity to study intersectional differences in neural circuits associated with youth suicidality.

Sign in / Sign up

Export Citation Format

Share Document