Evelyn·Jul 31A case where log-rank misses a real differenceLast week I ran log-rank on a reconstructed trial’s curves for a meta-analysis. The two curves obviously differed. One arm ran higher for…
Evelyn·Jul 25KM curves in R: plotting, and pulling median with a CIIn previous posts I wrote about what the confidence interval on a median survival actually is. A range of times, not median ± error, and…
Evelyn·Jul 24Reading a published KM curve, and what the censoring section doesn’t say out loudEvery trial’s methods section has a sentence that reads more or less like this. Patients were censored at their last documented visit or at…
Evelyn·Jul 17Reconstructing a Surv() object from a published KM curveLast Friday I wrote about Surv(time, event) on a shared trial IPD. Four decisions the trial made, hidden in two columns, re-expressed in R…
chat_999·Apr 19大義名分と正当性☆me☆ 大義名分、正当性、って、面白いね（笑） 私は、コレを幻想、人間が後から制定したモノだ、と、思っていたけれど、、、 現実社会にて、キチンと働いているよ。個人でも、集合体でも、この枠組み、キチンと働いている。
