Selected recent work

(2020). Errors in the statistical justification for the "credibility ceiling" remain uncorrected. Journal of Clinical Epidemiology. In press.

Preprint

(2020). Fundamental problems with the "credibility ceiling" method for meta-analyses. Journal of Clinical Epidemiology. In press.

Preprint

(2020). New statistical metrics for multisite replication projects. Journal of the Royal Statistical Society: Series A. In press.

Preprint

(2020). Robust metrics and sensitivity analyses for meta-analyses of heterogeneous effects. Epidemiology.

PDF

(2020). Evidence relating healthcare provider burnout and quality of care. Annals of Internal Medicine. [Critique of the excess significance test method for meta-analyses].

PDF

(2020). Reducing meat consumption by appealing to animal welfare: Protocol for a meta-analysis and theoretical review. Systematic Reviews.

PDF

(2020). A simple, interpretable conversion from Pearson’s correlation to Cohen’s d for continuous exposures. Epidemiology.

PDF

(2019). Estimating publication bias in meta-analyses: A meta-meta-analysis across disciplines and journal tiers. Under review.

Preprint

(2019). Challenges and suggestions for defining replication "success" when effects may be heterogeneous: Comment on Hedges and Schauer (2019). Psychological Methods.

PDF

(2019). Sensitivity analysis for publication bias in meta-analyses. Under review.

Preprint

Software

*
All R Stata GUIs Statistics Psychology

Regression-based causal mediation analysis with a treatment-mediator interaction term.

Software in R, Javascript, and Qualtrics to design and analyze psychology experiments that use fine-grained mouse-tracking measures of perceptual category competition.

Conducts sensitivity analysis for publication bias in meta-analyses.

Contains functions to estimate the proportion of effects stronger than a threshold of scientific importance, to make effect size conversions, and to compute and format estimates and inference.

Conducts sensitivity analysis for unmeasured confounding for observational studies and meta-analysis, and conducts analogous sensitivity analyses for selection bias.

Calculates the E-value, a measure of sensitivity to unmeasured confounding.

Calculates the E-value, a measure of sensitivity to unmeasured confounding.

Computes metrics of outcome-wide evidence strength for studies testing multiple correlated outcomes.

Conducts statistical analyses for multisite replication projects.

Simulates a longitudinal dataset with time-varying covariates with user-specified correlation structure.