Regression-based Test Research Articles

Seroprevalence of infectious bursal disease and its associated risk factors in Ethiopia: a systematic review and meta-analysis

ABSTRACT The poultry population is an integral part of Ethiopia’s Gross Domestic Product (GDP) but, due to various infectious diseases such as infectious bursal disease (IBD), the expected economic impact in the country remains limited. The status of the disease in Ethiopia is obscured; thus, a systematic review and meta-analysis were employed to estimate the overall pooled prevalence of IBD in Ethiopia. Meta-analysis was conducted to determine the effects of each identified risk factor, while meta-regression and sub-group analysis were employed to assess the relationship between study-level covariates and effect size. The pooled prevalence of IBD in Ethiopia was 69.4% (95% CI 30.7–96.2), while the pooled logit prevalence was 0.94 (95% CI: 0.68–1.20) with significant inter-study variance (Q test = 948.28, df = 43, P < 0.001; τ 2 = 0.71, I 2 = 95.47%). A small-study effect was detected in the regression-based Egger test (Prob > |z| < 0.0001). Significant variation was observed among different groups such as sex, age, breed, and type of farm of the chickens. The effect size for the study period from 2018 to 2021 was significantly lower by −0.204 compared to the study period from 2009 to 2015 (P < 0.0001. In conclusion, the IBD pooled prevalence estimate is high, even though the number of studies in the country is insufficient. The high prevalence of the disease requires prompt attention from all stakeholders in the sector to bring it under control through comprehensive disease prevention and control intervention strategies.

The market timing ability of bond mutual funds

We apply the nonparametric methodology of Jiang (Journal of Empirical Finance 10:399–425, 2003) to examine whether bond mutual funds can time the bond market by adjusting their portfolios' market exposure based on anticipated market movement. This approach offers several advantages over the widely used regression-based tests such as Treynor and Mazuy (Harvard Business Review 44(4):131–136, 1966) and Henriksson and Merton (The Journal of Business 54(4):513–533, 1981). In a comprehensive study covering the USA, UK, and China, we find some evidence of positive market timing of bond funds at the individual fund level. On average, bond funds show neutral to slightly negative market timing abilities. After controlling for public information, we find that a smaller number of bond funds successfully time the market based on private timing signals. In terms of categories, we find strong evidence of positive market timing for Government bond funds as a group, consistent with the findings of Huang and Wang (Management Science 60:2091–2109, 2014).

Evaluating USDA’s farm balance sheet forecasts

PurposeThe United States Department of Agriculture Farm Balance Sheet forecasts provide important, timely information on the financial assets and debt in the U.S. farm sector. Despite their prominent role in policy and decision making, the forecasts have not been rigorously evaluated. This research examines the degree to which the USDA’s Farm Balance Forecasts are optimal predictors of subsequent official estimates.Design/methodology/approachFollowing prior studies of USDA’s farm income forecasts, archived asset and debt forecasts from 1986 through 2021 are used in regression-based tests of bias and efficiency.FindingsForecasts from 1986–2021 are found to be unbiased but inefficient. The forecasts have a tendency to over-react to new information early in the revision process.Originality/valueThese findings can be helpful for forecast users in adjusting their expectations and for forecasters in adjusting the current forecasting methods.

Long-term health effects of outdoor air pollution on asthma and respiratory symptoms among adults in low-and middle-income countries (LMICs): a systematic review and meta-analysis

BackgroundFew studies have investigated the association between long-term exposure to outdoor air pollution and the risk of asthma and respiratory symptoms in low-and middle-income countries (LMICs).ObjectiveTo systematically evaluate the epidemiological evidence regarding the associations between long-term exposure to outdoor air pollution and respiratory symptoms in LMICs.MethodsWe searched for literature published between January 1946 and September 2022 in Embase (Ovid), Medline (Ovid), and Web of Science (Core Collection). The air and gaseous pollutants studied included particulate matter (PM2.5 and PM10), nitrogen dioxide (NO2), and black carbon (BC), and exposure was 1-year duration or more. We conducted a systematic review and meta-analysis with a random-effects model to calculate the relative risk (RR) estimates. A rerun of the databases was conducted in November 2023 with no eligible studies found.ResultsOf the 1,346 studies identified, only six met our inclusion criteria, and these six reported PM2.5, PM10, and NO2 with asthma as the main outcome. Three of these included studies were further included in the meta-analysis because they had data on the same exposure and outcome (PM2.5 and asthma). The main result of our study showed a borderline significant association between a 10 µg/m3 increase in exposure to PM2.5 and an increased risk of asthma (RR 1.21, 95% CI 0.93, 1.50). There was evidence of considerable heterogeneity (I2 = 75.87%). The regression-based Egger test for small-study effects showed no significant publication bias among these three studies.ConclusionOur results indicate that long-term exposure to PM2.5 increases the risk of asthma in LMICs, but studies are scarce and there is a large need for more research in LMICs in this field.Systematic Review Registrationhttps://journals.plos.org/plosone/article?id=10.1371/journal.pone.0288667, PROSPERO (CRD42022311326).

Prenatal PFAS exposure, gut microbiota dysbiosis, and neurobehavioral development in childhood

Studies on the role of the gut microbiota in the associations between per- and polyfluoroalkyl substance (PFAS) exposure and adverse neurodevelopment are limited. Umbilical cord serum and faeces samples were collected from children, and the Strengths and Difficulties Questionnaire (SDQ) was conducted. Generalized linear models, linear mixed-effects models, multivariate analysis by linear models and microbiome regression-based kernel association tests were used to evaluate the associations among PFAS exposure, the gut microbiota, and neurobehavioural development. Perfluorohexane sulfonic acid (PFHxS) exposure was associated with increased scores for conduct problems and externalizing problems, as well as altered gut microbiota alpha and beta diversity. PFHxS concentrations were associated with higher relative abundances of Enterococcus spp. but lower relative abundances of several short-chain fatty acid-producing genera (e.g., Ruminococcus gauvreauii group spp.). PFHxS exposure was also associated with increased oxidative phosphorylation. Alpha and beta diversity were found significantly associated with conduct problems and externalizing problems. Ruminococcus gauvreauii group spp. abundance was positively correlated with prosocial behavior scores. Increased alpha diversity played a mediating role in the associations of PFHxS exposure with conduct problems. Our results suggest that the gut microbiota might play an important role in PFAS neurotoxicity, which may have implications for PFAS control.

Are they worth it? – An evaluation of predictions for NBA ‘Fantasy Sports’

Abstract‘Fantasy Sports’ - an internet-based game in which participants chose virtual teams of real professional athletes - has recently gained in popularity. Various firms provide projections regarding athletes’ future performance to help participants choose their virtual teams. We evaluate such forecasts based on 1658 projections regarding NBA basketball of four selected projection providers that were collected in February 2022. We calculate standard measures of forecast quality and find that the use of professional forecasts reduces the errors made in naïve forecasts, but only to a moderate extent. Applying regression-based tests of forecast efficiency, we find that the predictions are inefficient and, in some cases, even biased. Third, pairwise comparisons of the accuracy of the providers suggest notable differences among such providers in the short run. We use a simple optimization algorithm to choose a virtual team for each match day and feed it with the forecasts of the providers. Subsequently, we rank the providers according to the score obtained by these teams. We find small, although in one case significant, long-run differences between the providers, among whom each provides better accuracy than that of a naïve projection based on these athletes’ past performances. Finally, we simulate one-on-one competition among various forecast providers to ascertain the long-term profitability of their services. Given the small magnitude of the detected differences, our results, in brief, raise doubts as to whether the forecasts provided are worth the money.

Effects of exergaming on executive functions of children: a systematic review and meta-analysis from 2010 to 2023

BackgroundExecutive function plays a crucial role in children’s cognitive development, academic performance, as well as their physical and mental health. This study aims to assess the impact of exergaming on executive functions in pediatric populations.MethodsThe criteria of inclusion were randomized controlled trials of exergaming intervention and evaluation of executive function in children aged 4–12 years. A meta-analysis was performed in databases of China National Knowledge Infrastructure (in Chinese), Wan Fang (in Chinese), Web of Science, Embase, and PubMed, from January 2010 to February 2023, following the PRISMA guidelines. Risk of bias was assessed by the Jadad scale, the Cochrane risk of bias assessment tool, funnel plot, and regression-based Egger test. The Review Manager 5.3 was used to analyze the included articles using a random-effects model, and the effects were calculated as standardized mean difference (SMD).ResultsEleven experimental studies with children (n = 508) were included. Exergaming was found to have a positive impact on children’s cognitive flexibility (SMD = 0.34, 95%CI [0.17,0.52], P < 0.01), inhibitory control (SMD = 0.57, 95%CI [0.31,0.83], P < 0.01), and working memory (SMD = 0.26, 95%CI [0.02,0.51], P < 0.05). The publication bias were observed.ConclusionsExergaming has the potential to improve executive functions in children. More studies with rigorous designs are warranted to explore the specific effects of exergaming intervention. This study was registered on the PROSPERO (CRD42023401526).

The Influence of Electroconvulsive Therapy (ECT) on Brain-Derived Neurotrophic Factor (BDNF) Plasma Level in Patients with Schizophrenia-A Systematic Review and Meta-Analysis.

The main aim of this systematic review and meta-analysis is to establish whether there is a correlation between the brain-derived neurotrophic factor (BDNF) level and electroconvulsive therapy (ECT) treatment and the reduction in psychotic symptoms in patients diagnosed with schizophrenia. A systematic search of PubMed/Medline, Cochrane Library, Web of Science, Scopus and Embase was conducted up to March 2023. Inclusion criteria: studies in which adult patients with schizophrenia treated with antipsychotic medication received ECT therapy and had the BDNF level measured before and after ECT treatment. Exclusion criteria: animal and in vitro studies or studies not involving complete information about the treatment and concentration of BDNF in plasma. The risk of bias was assessed using Egger's regression-based test for meta-analysis with continuous outcomes. Six studies comprising 248 individuals with schizophrenia were included. A statistically significant increase in BDNF levels after ECT treatment was observed only in two studies (p < 0.001 and p < 0.027, respectively), whereas in four other studies, an upward trend without statistical significance was noticed. The estimated overall size effect revealed that ECT therapy caused a slight change in the BDNF level but without statistical significance (ES = -0.328). Different numbers of ECT procedures (4-10), final measurement of the BDNF level made at a different time point, using bilateral or unilateral electrode positioning during ECT and treatment with different combinations of typical or atypical antipsychotic medications may be potential reasons for the lack of statistical significance in the changes in BDNF levels after treatment. Data regarding the measurement of BDNF levels pre and post ECT therapy in patients with schizophrenia are very limited without an extended follow-up period and evaluation of mental health change. Our meta-analysis showed that treatment with ECT therapy and antipsychotic medication increases serum BDNF levels in patients with drug-resistant schizophrenia compared to patients treated with medication only; however, this effect is not statistically significant.

Beyond compliance: A randomized trial of DEI statements and subsequent signals for job seekers with disabilities

BackgroundCompanies are increasingly attending to their Diversity, Equity, and Inclusion (DEI) statements in their public-facing documents, including job advertisements. Little is known about how various decisions about language type or placement within a job ad are received by marginalized populations, including those living with disabilities. ObjectiveThis study aims to explore various options for DEI statements within job ads, to assess whether these choices have a measurable impact on potential job seekers living with disabilities. MethodsAn experiment with a 2 × 2 × 2 design was used which varied the DEI language (legal/traditional or heartfelt), the placement at the top or the bottom of the job advertisement, and whether the company offered the opportunity to indicate a need for accommodations even before the interview. Quantitative and qualitative items measured general reactions to the statement, intent to apply to this company, willingness to disclose a need for accommodations, how supportive the company seemed, and impressions of the company's level of trustworthiness. Statistical tests include t-tests, z-tests for proportions, and regression-based tests for mediation. ResultsHeartfelt statements and upfront placement inspired more positive reactions from participants, with trust mediating these positive relationships. However, participants also expressed skepticism that, regardless of language, bias would still exist in the reality of the job. ConclusionsThese results suggest that job candidates with disabilities are not indifferent to a company's DEI framing and prefer heartfelt language and earlier placement, though statements without direct evidence can raise suspicion about whether the actions would stay true to the words.

Conceptual and analytical approaches for modelling the developmental origins of inequality.

In many species, individuals that experience harsh conditions during development have poor health and fitness outcomes in adulthood, compared with peers that do not. These early-life contributions to inequality are often attributed to two classes of evolutionary hypotheses: Developmental Constraints (DC) models, which focus on the deleterious effects of low-quality early-life environments, and Predictive Adaptive Response (PAR) hypotheses, which emphasize the costs individuals incur when they make incorrect predictions about conditions in adulthood. Testing these hypotheses empirically is difficult for conceptual and analytical reasons. Here, we help resolve some of these difficulties by providing mathematical definitions for DC, PAR (particularly focusing on 'external' PAR) and related concepts. We propose a novel, quadratic regression-based statistical test derived from these definitions. Our simulations show that this approach markedly improves the ability to discriminate between DC and PAR hypotheses relative to the status quo approach, which uses interaction effects. Simulated data indicate that the interaction effects approach often conflates PAR with DC, while the quadratic regression approach yields high sensitivity and specificity for detecting PAR. Our results highlight the value of linking verbal and visual models to a formal mathematical treatment for understanding the developmental origins of inequitable adult outcomes. This article is part of the theme issue 'Evolutionary ecology of inequality'.

Differentially Private Nonlinear Causal Discovery from Numerical Data

Recently, several methods such as private ANM, EM-PC and Priv-PC have been proposed to perform differentially private causal discovery in various scenarios including bivariate, multivariate Gaussian and categorical cases. However, there is little effort on how to conduct private nonlinear causal discovery from numerical data. This work tries to challenge this problem. To this end, we propose a method to infer nonlinear causal relations from observed numerical data by using regression-based conditional independence test (RCIT) that consists of kernel ridge regression (KRR) and Hilbert-Schmidt independence criterion (HSIC) with permutation approximation. Sensitivity analysis for RCIT is given and a private constraint-based causal discovery framework with differential privacy guarantee is developed. Extensive simulations and real-world experiments for both conditional independence test and causal discovery are conducted, which show that our method is effective in handling nonlinear numerical cases and easy to implement. The source code of our method and data are available at https://github.com/Causality-Inference/PCD.

A systematic review and meta-analysis of antimicrobial resistance knowledge, attitudes, and practices: Current evidence to build a strong national antimicrobial drug resistance narrative in Ethiopia.

Antimicrobial resistance (AMR) is a silent pandemic that has claimed millions of lives, and resulted in long-term disabilities, limited treatment options, and high economic costs associated with the healthcare burden. Given the rising prevalence of AMR, which is expected to pose a challenge to current empirical antibiotic treatment strategies, we sought to summarize the available data on knowledge, attitudes, and practices regarding AMR in Ethiopia. Articles were searched in international electronic databases. Microsoft Excel spreadsheet and STATA software version 16 were used for data extraction and analysis, respectively. The Preferred Reporting Items for Systematic Reviews and Meta-Analysis 2020 checklist was followed. The methodological quality of the studies included was assessed by the Joana Briggs Institute critical appraisal checklists. The random-effect meta-analysis model was used to estimate Der Simonian-Laird's pooled effect. Statistical heterogeneity of the meta-analysis was checked through Higgins and Thompson's I2 statistics and Cochran's Q test. Publication bias was investigated by funnel plots, and the regression-based test of Egger for small study effects with a P value < 0.05 was considered to indicate potential reporting bias. In addition, sensitivity and subgroup meta-analyses were performed. Fourteen studies with a total of 4476 participants met the inclusion criteria. Overall, the pooled prevalence of good AMR knowledge was 51.53% [(95% confidence interval (CI): 37.85, 65.21), I2 = 99.0%, P <0.001]. The pooled prevalence of favorable attitudes and good practices were 63.43% [(95% CI: 42.66, 84.20), I2 = 99.6, P <0.001], and 48.85% [(95% CI: 38.68, 59.01), I2 = 93.1, P <0.001] respectively. In conclusion, there is a significant knowledge and practice gap on AMR among the general public, patients, and livestock producers. As a result, we call for greater educational interventions to raise awareness and build a strong national AMR narrative.

Ethiopians' knowledge of and attitudes toward epilepsy: A systematic review and meta-analysis.

Epilepsy remains one of the world's most common neurological diseases, but it appears to be widely misunderstood, particularly in under-resourced countries like Ethiopia. Improving individuals' knowledge and attitude toward epilepsy is critical for reducing the multifaceted impacts of epilepsy. Therefore, in this study, we sought to estimate the pooled levels of good knowledge and a favorable attitude toward epilepsy and also identify the associated factors using available data collected from different segments of the population. Articles were searched in international electronic databases. A standardized Microsoft Excel spreadsheet and STATA software version 16 were used for data extraction and analysis, respectively. The Preferred Reporting Items for Systematic Reviews and Meta-Analysis (PRISMA) checklist was used to write this report. The random-effect meta-analysis model was used to estimate Der Simonian-Laird's pooled effect. Statistical heterogeneity of the meta-analysis was checked via Higgins and Thompson's I 2 statistics (0-100%), and Cochran's Q test at P < 0.10. Subgroups, based on the study regions, and sensitivity analyses were also performed. Publication bias was examined subjectively using funnel plots and objectively using the nonparametric rank correlation test of Begg and the regression-based test of Egger for small study effects with P < 0.05 considered to indicate potential publication bias. Furthermore, the Trim-and-fill method of Duval and Tweedie was used to explore sources of publication bias for the favorable level of attitudes toward epilepsy. A total of 12 studies with 6,373 study participants and 10 studies with 5,336 study participants were included to estimate the pooled level of good epilepsy knowledge and favorable attitudes respectively. The overall estimated levels of good epilepsy knowledge and favorable attitudes toward epilepsy among Ethiopians were 47.37% [(95% CI: 35.00, 59.74), I 2 = 99.2, P < 0.001] and 46.83%[(95% CI: 32.75, 60.90), I 2 = 99.2, P < 0.001] respectively. Subgroup analysis revealed that the pooled level of good epilepsy knowledge was 48.51% [(95% CI: 38.95, 58.06), I 2 = 95.6%, P < 0.001] in the Amhara region. In the current review, we found out that there is a huge knowledge gap and an unfavorable level of attitudes towardepilepsy, which demand immediate public health action as well as a targeted policy intervention.

Prevalence and determinants of opportunistic infections among HIV-infected adults receiving antiretroviral therapy in Ethiopia: A systematic review and meta-analysis.

Reliable data on the burden of opportunistic infections (OIs) after the initiation of antiretroviral therapy (ART) is critical for planning health services and reducing OI-related morbidity and mortality. Nevertheless, there has been no nationally representative information on the prevalence of OIs in our country. Therefore, we have undertaken this comprehensive systematic review and meta-analysis to estimate the pooled prevalence, and identify factors associated with the development of OIs in Human Immunodeficiency Virus (HIV)-infected adults receiving ART in Ethiopia. Articles were searched in international electronic databases. A standardized Microsoft Excel spreadsheet and STATA software version 16 were used for data extraction and analysis, respectively. The Preferred Reporting Items for Systematic Reviews and Meta-Analysis (PRISMA) checklist was used to write this report. The random-effect meta-analysis model was used to estimate the pooled effect. The statistical heterogeneity of the meta-analysis was checked. Subgroup and sensitivity analyses were also performed. Publication bias was examined in funnel plots and the nonparametric rank correlation test of Begg and the regression-based test of Egger. Association was expressed through a pooled odds ratio (OR) with a 95% Confidence Interval (CI). A total of 12 studies with 6,163 study participants were included. The pooled prevalence of OIs was 43.97% [95% CI (38.59, 49.34)]. Poor adherence to ART [OR, 5.90, 95% CI (3.05, 11.40)], under nutrition [OR, 3.70, 95% CI (2.01, 6.80)], CD4 T lymphocyte count <200 cells /μL [OR, 3.23 95% CI (2.06, 5.07)], and advanced World Health Organization (WHO) HIV clinical stages [OR, 4.84 95% CI (1.83, 12.82)] were determinants of OIs. The pooled prevalence of OIs among adults taking ART is high. Poor adherence to ART, under nutrition, a CD4 T lymphocyte count <200 cells /μL, and advanced WHO HIV clinical stages were factors associated with the development of OIs.

Inflammatory cell response following ST-elevation myocardial infarction treated with primary percutaneous coronary intervention and its impact on cardiovascular outcomes: A systematic review and meta-analysis

BackgroundInflammatory responses post STEMI may mediate major adverse cardiovascular events (MACE). This is the first systematic review to map leukocyte response following a STEMI and its association with outcomes. MethodsWe systematically searched EMBASE and Medline for studies of STEMIs undergoing primary PCI. Eligible studies reported leukocytes or its subtype plus either 30-day and/or 1-year MACE. Random effects model for pooled proportions was used to estimate 30-day and 1-year mortality and MACE. Meta-regression was used to estimate the effect of leukocyte counts on cardiovascular outcomes. Publication bias was assessed using Egger’s regression-based test. The review was registered with PROSPERO (CRD42019124991). ResultsOf the 3,813 studies meeting the preliminary search criteria, 24 cohort studies were eligible for inclusion, representing 19,074 persons [76.4% male (n = 14,539); mean age 61.6 years]. Leukocytes had a mean of 10.5x109 (SD 4.7) on admission and 11.1x109 (SD 3.3) at day one post STEMI. Neutrophils increased day one post STEMI, while lymphocytes decreased. There was limited data on other leukocyte subtypes and beyond day one. Estimated 30-day and 1-year all-cause mortality were 6.5% (95% CI 4.8-8.2, p <0.001) and 9.7% (95% CI 5.6-13.8, p <0.001), while the estimated 30-day and 1-year MACE were 14.9% (95% CI 5.3-24.4, p < 0.001) and 15.2% (95% CI 7.2-23.2, p < 0.001). The meta-analysis was limited by a high degree of heterogeneity between studies. ConclusionsThis review highlights the urgent need to better characterise inflammation post STEMI to identify mediators for the persistently high mortality and morbidity associated with STEMI.

Cyberostracism and knowledge sharing: The mediating role of social anxiety in e-learning environments

Ostracized individuals face a series of negative situations and feel social anxiety during their interactions with others. However, individuals with social anxiety may not share their knowledge because of the discomfort they experience in interactions. The aim of this study is to examine the mediating role of social anxiety in learner–learner interaction in terms of the relationship between cyberostracism and knowledge sharing in e-learning environments. A total of 268 pre-service teachers participated in the study. The study determined the relationship between the variables with Pearson’s correlation coefficients and used a regression-based mediation test to examine the mediator model. It found that cyberostracism was positively related to social anxiety in learner–learner interactions in e-learning environments and negatively associated with knowledge sharing. There was also a negative relationship between social anxiety and knowledge sharing. Moreover, cyberostracism predicted knowledge sharing in e-learning environments. Finally, we determined that social anxiety in learner–learner interactions was the full mediator in the relationship between cyberostracism and knowledge sharing. Therefore, it can be argued that cyberostracism is not a direct predictor of pre-service teachers’ knowledge sharing behavior in e-learning environments and pre-service teachers avoid knowledge sharing due to social anxiety caused by cyberostracism. The research results also led to suggestions for practice and research.

MultiMiAT: an optimal microbiome-based association test for multicategory phenotypes.

Microbes can affect the metabolism and immunity of human body incessantly, and the dysbiosis of human microbiome drives not only the occurrence but also the progression of disease (i.e. multiple statuses of disease). Recently, microbiome-based association tests have been widely developed to detect the association between the microbiome and host phenotype. However, the existing methods have not achieved satisfactory performance in testing the association between the microbiome and ordinal/nominal multicategory phenotypes (e.g. disease severity and tumor subtype). In this paper, we propose an optimal microbiome-based association test for multicategory phenotypes, namely, multiMiAT. Specifically, under the multinomial logit model framework, we first introduce a microbiome regression-based kernel association test for multicategory phenotypes (multiMiRKAT). As a data-driven optimal test, multiMiAT then integrates multiMiRKAT, score test and MiRKAT-MC to maintain excellent performance in diverse association patterns. Massive simulation experiments prove the success of our method. Furthermore, multiMiAT is also applied to real microbiome data experiments to detect the association between the gut microbiome and clinical statuses of colorectal cancer as well as for diverse statuses of Clostridium difficile infections.

Causal Gene Identification Using Non-Linear Regression-Based Independence Tests.

With the development of biomedical techniques in the past decades, causal gene identification has become one of the most promising applications in human genome-based business, which can help doctors to evaluate the risk of certain genetic diseases and provide further treatment recommendations for potential patients. When no controlled experiments can be applied, machine learning techniques like causal inference-based methods are generally used to identify causal genes. Unfortunately, most of the existing methods detect disease-related genes by ranking-based strategies or feature selection techniques, which generally return a superset of the corresponding real causal genes. There are also some causal inference-based methods that can identify a part of real causal genes from those supersets, but they are just able to return a few causal genes. This is contrary to our knowledge, as many results from controlled experiments have demonstrated that a certain disease, especially cancer, is usually related to dozens or hundreds of genes. In this work, we present an effective approach for identifying causal genes from gene expression data by using a new search strategy based on non-linear regression-based independence tests, which is able to greatly reduce the search space, and simultaneously establish the causal relationships from the candidate genes to the disease variable. Extensive experiments on real-world cancer datasets show that our method is superior to the existing causal inference-based methods in three aspects: 1) our method can identify dozens of causal genes, and 1/3 ∼1/2 of the discovered causal genes can be verified by existing works that they are really directly related to the corresponding disease; 2) The discovered causal genes are able to distinguish the status or disease subtype of the target patient; 3) Most of the discovered causal genes are closely relevant to the disease variable.

A Regression-Based Test of Shaped Consumption Utility Functions

A Regression-Based Test of Shaped Consumption Utility Functions

Common permutation methods in animal social network analysis do not control for non-independence

The non-independence of social network data is a cause for concern among behavioural ecologists conducting social network analysis. This has led to the adoption of several permutation-based methods for testing common hypotheses. One of the most common types of analysis is nodal regression, where the relationships between node-level network metrics and nodal covariates are analysed using a permutation technique known as node-label permutations. We show that, contrary to accepted wisdom, node-label permutations do not automatically account for the non-independences assumed to exist in network data, because regression-based permutation tests still assume exchangeability of residuals. The same assumption also applies to the quadratic assignment procedure (QAP), a permutation-based method often used for conducting dyadic regression. We highlight that node-label permutations produce the same p-values as equivalent parametric regression models, but that in the presence of non-independence, parametric regression models can also produce accurate effect size estimates. We also note that QAP only controls for a specific type of non-independence between edges that are connected to the same nodes, and that appropriate parametric regression models are also able to account for this type of non-independence. Based on this, we suggest that standard parametric models could be used in the place of permutation-based methods. Moving away from permutation-based methods could have several benefits, including reducing over-reliance on p-values, generating more reliable effect size estimates, and facilitating the adoption of causal inference methods and alternative types of statistical analysis.