Author statement
The conception and design of the study were completed by B.D. and I.H.
Acquisition of data was done by B.D. and M.R. Analysis of the data was
done by B.D., I.H and M.R. Interpretation of data was handled by B.D.,
I.H and M.R. The manuscript was completed by B.D. with critical
revisions done by all authors. All authors gave final approval of the
completed manuscript.
Figure 1 : The likelihood that testing in further RCTs will not
be required by the regulatory agencies reflects the Weber-Fechner law:
the higher the signal [effect size, here expressed as logarithm base
10 of odds ratio, [log(OR)], the greater the probability of the
treatment approval without requirement for conducting confirmatory RCT.
Figure 2: 2a) a relation between recognition validity(expressed as positive predictive value, PPV) and d’ [i.e., the
standardized distance between distributions of the signal (i.e., no
further RCTs needed and noise (further RCTs required)]; 2b) a
relationship between d’ and “recognition heuristic”effect size threshold value . Optimal cut-off is about log(OR) of
1; 2c) the largest evidentiary support is found for effect size of 1.11
at moderate evidentiary support p=0.033. p values for larger thresholds
have not reached traditional statistical evidence at p<0.05.
(Fig 2c).