回帰分析のイントロ講義で”one unit increase in ~ is associated with-“みたいな解釈を教えることが多いと思うけど、これって学生を混乱させるだけなんじゃないかなって気がしている。最初からあくまで条件付き期待値の比較であるとした方が、例えばダミー変数の考え方もすんなり入るんじゃないかな。
#J_Epidemi
Advanced Publication:
When and how to split the follow-up time in the analysis of epidemiological or clinical studies with follow-ups
Masao Iwagami et al.
@J_Epidemi
@she_knows_a_key
#J_Epidemi
Editors' Picks:
Exchangeability of Measures of Association Before and After Exposure Status Is Flipped: Its Relationship With Confounding in the Counterfactual Model
Etsuji Suzuki et al.
@J_Epidemi
あの2000年の経済学賞(McFaddenとHeckman)後のIBCでの会長講演録。受賞成果とパラレルに展開してきた疫学・生物統計分野の足跡を紹介。この手の賞の行方を眺めるのに、20年経っても一読の価値があります
Are statistical contributions to medicine undervalued?
Our new pedagogic paper with a fresh numerical example for longitudinal causal inference is published as a part of “Pitfalls and Tips” series in Journal of Epidemiology
Understanding marginal structural models for time-varying exposures: pitfalls and tips
Our new pedagogic paper with a fresh numerical example for longitudinal causal inference is published as a part of “Pitfalls and Tips” series in Journal of Epidemiology
Understanding marginal structural models for time-varying exposures: pitfalls and tips
1st piece of new methodological review series in Journal of Epidemiology. Rather than providing introductory tutorial for causal DAGs, we reviewed technical difficulties often overlooked when learning them.
Suzuki et al. Causal diagrams: pitfalls and tips
Our new pedagogic paper with a fresh numerical example for longitudinal causal inference is published as a part of “Pitfalls and Tips” series in Journal of Epidemiology
Understanding marginal structural models for time-varying exposures: pitfalls and tips
immortal time biasは結果で結果を説明するような、凡そ考えうるバイアスでも最も間抜けなものですが、バイアスの「構造」を説明するのは難しい。このeditorialの良いところはimmortal timeの扱い方によりバイアスを分類している点
Causal diagrams for immortal time bias
New paper: To obtain interval estimates of prediction measures like C-stat, bootstrapping can (& should) be used for bootstrap-based optimism-corrected estimates. This 2-stage (or "nested loop") bootstrap resampling is computer-intensive but works well.
Our invited commentary on the ASA Biopharma Section-Working Group's series of papers is just out🤘
Comment on “Biostatistical Considerations When Using RWD and RWE in Clinical Studies for Regulatory Purposes: A Landscape Assessment”
Our new paper led by Dr. Hagiwara proposes to conduct ITT analysis more formally by borrowing the concept of stochastic dynamic treatment regimes. ITT analysis cannot provide "effectiveness" nor "treatment policy effect" estimates by itself.
Tutorial in Biostatistics
Statistical plasmode simulations–Potentials, challenges and recommendations. Nicholas Schreck, Alla Slynko, Maral Saadati, Axel Benner. Statistics in Medicine.
RCTのメタアナリシスに記載されるCochraneのRisk of Bias評価はかねて有名無実と思っていたけど、この改訂版は見た目だけ残して観察研究用のROBINS-Iと整合性を図るため全面刷新という感じで驚いた
RoB 2: a revised tool for assessing risk of bias in randomised trials