There is a crucial mistake in interpreting the Hypothesis testing – What means? if P>0.05.
Hinman said :”in……chronic knee pain, neither laser nor needle acupuncture conferred benefit over sham for pain or function (Dr. Fan notes: Her statement was based on P>0.05). Our findings do not support acupuncture for these patients”
From the perspective of hypothesis testing in Statistics, if acupuncture has better results and with significant difference over the primary control (no-treatment group), p<0.05, we can conclude that “acupuncture is effective”- no matter what the result get from the comparing to the secondary control, such as “sham laser acupuncture”, but Hinman intentionally does not report this effectiveness in her conclusion; if acupuncture has better results over “laser acupuncture” and “sham laser acupuncture”, without significant in statistics, p>0.05, we can conclude that “acupuncture is better than the laser acupuncture, and sham laser acupuncture, but need more studies to confirm”. We can’t conclude that “acupuncture is not effective” because that there are no significant difference in statistics between acupuncture and “laser acupuncture”, or between acupuncture and “sham acupuncture” does not mean there is no difference between these treatments clinically. Hinman et al mis-interpreter the results and violates the basic principle of Statistics.