Estimands is the simplest concept with the most difficult implementation. With the release of the final ICH Guideline in Estimands today, this first of several blogs to follow delves into the topic.
No. 14: Estimands – WHAT Is It All About?
No. 13: Unconsciously Biased and Consciously Unbiased
Implicit models in the back of our minds can creep into explicit models creating biased predictions that have societal implications.
No. 12: Models – Implicit and Explicit
If we fail to acknowledge that we have biases and assumptions that influence our assessment of 'objective facts,' then we delude ourselves. Our perception of reality and how we judge evidence is colored by our beliefs which arise from our specific experiences.
No. 11: Some Beliefs in Priors
The probability that the null hypothesis is true is 0.50. How should we interpret that and then write it down mathematically?
No. 10 – Always do Subgroup IDENTIFICATION
You may have heard, “Always do subgroup analysis, but never believe them.” Don't believe this.
No. 9: Case Study – Genetic Subgroups and CV Disease
The over-reliance on p-values can lead to misinterpretation of data and a $150 million bet on a subgroup with scant evidence.
No. 8: Let’s Get Real – Bayes and Biomarkers
How do we know when an observed effect is real or spurious?
No. 7: What does p<0.05 mean anyway?
Some people say, "A p-value=0.05 is not very much evidence against the null hypothesis." Well then, how much evidence is it?
No. 6: Détente – The Peaceful Co-Existence of Significance Levels and Bayes
pr(B|A) ≠ 1 – pr(A|B). Why do we act like it ?!
No. 5: pr(You’re Bayesian) > 0.50
For too long statisticians have been peddling pr(data|hypothesis) when scientists [indeed all of us] want pr(hypothesis|data).
analytixthinking.blog 