Of course it's useless for her. It's based on the opinion of people you consider a useless source of information.
Because it's otherwise completely uninteresting.
Still, you don't answer the question. Do you actually have a superior strategy to studying the information available and making up your own mind when you don't feel you can trust the experts? I am beginning to think not. At least, not one you have sufficient confidence in to put up for public scrutiny.
I don't understand your reasoning here. The results of all the pre-approval trials combined showed a total of five cases out of 10,054 compared to one out of 4633 in those given placebo.
That's a rate of 0.0005 in the vaccinated population compared to a rate of .0002 in the unvaccinated population - an increase of approximately 150%. However, the rates are so low that large sample sizes are required to determine if the increase is statistically significant at the 0.05 level. It wasn't enough to say definitely that the vaccine was the cause, but it was enough to decide that further research was appropriate in order to make the determination with a high level of confidence.
BTW, I didn't catch the base population rate earlier. A quick computation gives an p-value of 0.0693 for 5 out of 10054 with a probability of 1 in 4633. While not significant at the 0.05 level, it would have been quite appropriate to take this result as an indicating that further investigation is needed.
You say a sample size of 1,000,000 is needed in order to make that finding statistically significant, then you say a sample size of 10,000(or less - it's confusing because you talk about a single trial, yet the results are from multiple trials) would establish a significant difference. Wouldn't it just lead to the same conclusion again?
I'm sorry about the confusing terminology. But the answer is not necessarily. Multiple experiments (which can consist of many experimental units - in this case an experimental unit would be a person who was vaccinated) can provide superior discrimination over a single experiment with more units. Consider the Million $ challenge. The prelim and the final test can both have a p-value of 0.001 because when multiplied together, they give a probability of 0.000001.
I know it seems contrary to intuition at first, but it's major reason why an iterated approach to experimentation with relatively small numbers of experimental units in each iteration is a much better approach than a single large experiment. You can actually gain more information from fewer experimental units with this sort of approach.
You might try thinking of it this way: if the original result was due to chance, then the another experiment is very unlikely to produce the same effect. On the other hand, if it isn't due to chance, the probability of seeing increased rates in the vaccine population is high.
But how could they possibly know that it was a problem when all they saw was something that occurred at a similar rate in the group given placebo, especially since it was biologically implausible?
An increase of 150% is NOT a similar rate. It wasn't a statistically significant increase because when you have such low rates it is extremely difficult to reject the null of no difference. Recall that the probability of a type II error (which is what happened in this case) is unrelated to the stated confidence level and can be much much higher.
What if it had been a similar slight difference in the number of car accidents or distribution of astrological signs?
Do they keep track of those factors because they are a possible consequence of the vaccine?
What kind of research are you involved in?
Do you disagree with any of the current recommendations of the ACIP? I asked first.
Yes, you've made it abundantly clear that we are all idiots being led by our noses.
No, no, no. You're taking offense at something inferred that was not meant to be implied. Such a strategy is usually skeptigirl's baliwick not yours. I have never said you were anything but biased. We are all biased, including me. I've explicitly said, more than once, that doesn't imply incompetence. It certainly doesn't imply stupidity. Why are you taking offense now? This comes across to me as yet another ploy to avoid answering my question regarding your personal opinion. The best defense is a good offense?
But in all honesty, I find it very hard to believe that the degree of distrust you exhibit would be influenced in the slightest by making changes in the process.
You're wrong about that. I would be quite satisfied if I felt the committee was composed of relatively few biased individuals or had a reasonable distribution of various different biases represented, thus achieving something like balance. It is the law. Is it really asking too much for a national policy setting committee to abide by the laws and regulations that govern it? I don't think so, but if it is, then the laws and regulations should be changed not ignored. It's been 8 years. What has changed? Anything?
If you're not going to be satisfied anyway (and I've seen no indication otherwise), why waste the effort?
Linda
Because we're having a friendly conversation. Because I've answered your questions to the best of my ability. Because I asked politely. Because it's relevant to the conversation. Why do you keep refusing? It can't take more time or effort to answer the question than it does to keep evading it.
If you're not going to give an answer, could you at least give me an honest reason why not?