New relevance and significance measures to replace p-values

Stahel, Werner A.

doi:10.1371/journal.pone.0252991

Cited by 13 publications

(8 citation statements)

References 19 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…This is achieved by an adaptive level α u which depends on the variance ratio c. The transformed p-value p = p 2 S then has exact linear T1E control and the corresponding p-value function produces a confidence region fully compatible with the skeptical p-value. The framework thus addresses an important point raised by Diggle (Diggle, 2020) about the need to accompany the skeptical p-value with suitable estimation procedures to assess the relevance of the observed effects (Stahel, 2021).…”

Section: Discussionmentioning

confidence: 99%

Beyond the two-trials rule: Type-I error control and sample size planning with the sceptical $p$-value

Held¹,

Micheloud²,

Balabdaoui³

2022

Preprint

View full text Add to dashboard Cite

We introduce a novel statistical framework to study replicability which simultaneously offers overall Type-I error control, an assessment of compatibility and a combined confidence region. The approach is based on a recently proposed reverse-Bayes method for the analysis of replication success. We show how the method can be recalibrated to obtain a family of combination tests for two studies with exact overall Type-I error control. The approach avoids the double dichotomization for significance of the two-trials rule and has larger project power to detect existing effects. It gives rise to a p-value function which can be used to compute a confidence region for the underlying true effect. If the effect estimates are compatible, the resulting confidence interval is similar to the meta-analytic one, but in the presence of conflict, the confidence region splits into two disjoint intervals. The proposed approach is applied to data from the Experimental Economics Replication Project.

show abstract

Section: Discussionmentioning

confidence: 99%

Beyond the two-trials rule: Type-I error control and sample size planning with the sceptical $p$-value

Held¹,

Micheloud²,

Balabdaoui³

2022

Preprint

View full text Add to dashboard Cite

show abstract

“…In the final step, a multivariable logistic regression analysis was conducted to identify the key risk factors of child stunting, the direction and size of their adjusted effects [ 36 ]. All the covariates that were significant at a 10% level of significance (p-values < 0.10) in the bivariate analyses were considered for the subsequent multivariable analysis [ 10 , 37 ]. There was no district level data clustering, thus a multivariable fixed effects logistic regression model was considered.…”

Section: Methodsmentioning

confidence: 99%

Child stunting and associated risk factors in selected food-insecure areas in Rwanda: an analytical cross-sectional study

Niragire¹,

Ndikumana²,

Nyirahabimana³

et al. 2022

Pan Afr Med J

View full text Add to dashboard Cite

Introduction stunting rates among the under-five children are generally high in Rwanda. They are unexpectedly lower than the national average stunting rate in some districts where poverty rates are the highest in the country. This study aimed to ascertain the key factors that protect children from stunting in these poorest areas, where stunting rates are lower than expected. Methods we analysed cross-sectional data from 2019/2020 Rwanda Demographic and Health Survey (RDHS) for 477 under-five children from Karongi, Rulindo, Nyanza, and Gisagara districts. Univariate and bivariate statistical analyses were used to find out the factors to retain for multivariable analysis. We obtained the key risk factors of child stunting through a multivariable binary logistic regression analysis. Results the child stunting rate in the study districts was 30 percent, which is three percent lower than the national average rate. Child stunting was negatively associated with a birth weight of at least 2.5 kg (AOR = 0.393, 95% CI: 0.180 - 0.856), a high household economic status (AOR = 0.506, 95%CI: 0.273 - 0.937), urban residence (AOR = 0.467; 95% CI: 0.222 - 0.984), and health insurance coverage (AOR = 0.418; 95% CI: 0.228 -0.767). Children aged one year and above, as well as female-headed households, were associated with at least three times and two times greater odds of child stunting than children aged below 12 months and those from male-headed households respectively. Conclusion the nutritional performance of children in the study districts is substantially driven by the high uptake of health insurance, which fosters increased access to healthcare services. To address child-stunting gaps in low-income areas in Rwanda, child nutrition programs should improve the utilization of healthcare services, and leverage the potential high effect of nutrition education, especially during pregnancy and lactation.

show abstract

“…Fortunately, the open science movement and shifts toward meta-science are helping bring these issues to light. Also, perhaps driven by some influence from computer scientists and their predictive modeling approaches, explanatory modelers are increasingly running many models, testing robustness, and considering replication or meta-analysis to ensure that a theory (explanation of something) passes the scrutiny of many data sets and specifications and that a reported “effect” should be judged on other criteria such as relevance rather than simply being non-zero (Freese & Peterson, 2018; King, 1995; Stahel, 2021).…”

Section: Understanding and Categorizing Computational Social Science Nowmentioning

confidence: 99%

Integrating Computer Prediction Methods in Social Science: A Comment on Hofman et al. (2021)

Breznau

2022

Social Science Computer Review

View full text Add to dashboard Cite

Machine learning and other computer-driven prediction models are one of the fastest growing trends in computational social science. These methods and approaches were developed in computer science and with different goals and epistemologies than those in social science. The most obvious difference being a focus on prediction versus explanation. Predictive modeling offers great potential for improving research and theory development, but its adoption poses some challenges and creates new problems. For this reason, Hofman et al. published recommendations for more effective integration of predictive modeling into social science. In this communication, I review their recommendations and expand on some additional concerns related to current practices and whether prediction can effectively serve the goals of most social scientists. Overall, I argue they provide a sound set of guidelines and a classification scheme that will serve those of us working in computational social science.

show abstract

New relevance and significance measures to replace p-values

Cited by 13 publications

References 19 publications

Beyond the two-trials rule: Type-I error control and sample size planning with the sceptical $p$-value

Beyond the two-trials rule: Type-I error control and sample size planning with the sceptical $p$-value

Child stunting and associated risk factors in selected food-insecure areas in Rwanda: an analytical cross-sectional study

Integrating Computer Prediction Methods in Social Science: A Comment on Hofman et al. (2021)

Contact Info

Product

Resources

About