Originally Posted by JamesGully
This is what you said:
"Skeptics can no longer say that homeopathic medicines are proven to be effective in treating a specific condition.
If anyone is asked, does Oscillococcinum have an effect on influenza or influenza-like syndrome, one must say YES."
I assumed the first sentence was a typo, otherwise we could all end the conversation in agreement. The meta-analysis does not support your claim that Oscillococcinum has an effect on influenza. It was qualified as "might".
Badly Shaved Monkey has provided a demonstration
that shows statistically significant differences can be expected most of the time if your initial group shows heterogeneity in the absence of any differences in treatment, such as special water
. I have also experienced this problem in actual practice (except that I recognized it was a problem I needed to deal with rather than manipulating it to my advantage to claim something that isn't true).
But now you have another problem. Here you are claiming that the effects of homeopathy are long term. That they don't show up in the short term is something to be expected. However, your fibrositis references all depend upon homeopathy having no long term effects
, as they are all cross-over studies. If you have any suspicion that the effects carry over from one treatment period to another, you cannot analyze them as though they do not.
So you face a choice. Either cling to your sepsis study as proof of homeopathy's effectiveness or your fibrositis studies, but they cannot both be true
The researchers performed a faulty analysis. When an appropriate analysis was performed, there were no statistically significant differences found.
I am surprised that you don't seem to be aware of this.
If you look at table 2 in this study
, you will see that there are no statistically significant differences between the two groups on any of the measures. The researchers elected to adjust the values for some of the differences in baseline measures, and it is the adjusted values that show statistically significant differences. That these two analyses lead to divergent results weakens your ability to draw conclusions from the data. And it invalidates any claims of certainty.
I don't have access to the full text of this study
. Until I see otherwise, I have to be suspicious that this study suffers from the same problems as the other two. And I find this line confusing - "...sniff alpha-1 and alpha-2 increases at 6 months versus baseline [did not correlate with] clinical outcomes in the active group." Are they saying that the EEG findings were unrelated to the clinical outcomes? Wouldn't that weaken the idea that the homeopathic treatment influenced either outcome?
Humans see pattern wherever they look, even when there is no connection to the elements. I think the problem is that while we recognize that the appearance of a pattern does not necessarily mean that there is a connection, many people have trouble letting that idea go. The "patterns" that you attribute to homeopathy are the same patterns that we expect to see from chance, bias (including cognitive), and a little fraud. There doesn't need to be a true treatment effect from homeopathy to generate the pattern that we see.
Two of these studies were outright negative. One study showed "significant" results that would be expected at least 50% of the time in the absence of treatment effects. That you consider this a "slam dunk" reflects very poorly on the veracity of your claims.