@sama@thorstenball I feel it no longer performs good at statistical analysis or script generation for it. Too many mistakes, too long scripts; overall, in statistical analysis it has poor understanding and is overcomplicated in execution.
@fopminui@dom_lucre Vaccines are to prevent diseases that would harm people both healthy and not. Vaccines don’t compete nor are replaceable with healthy habits.
@Retlouping Depends on the outcome; and in RCTs depend a if there is blinding. Likewise, blinding is a must for outcomes that can be subjective. But I agree that just because it is patient-reported it does not mean it is subjective.
There is no statistical solution to the problem of causal inference from observational data, it requires good scientific & statistical judgement, & knowledge of the literature. Folks like short cuts (propensity scores, E values etc) to avoid hard thinking. Short cuts don’t exist.