Probably the best we can do to compare $AVXL and $SAVA Cognitive efficacy distribution, but close to completely useless due trial design, endpoint and scoring as well as data presentation differences.
What we can see is that the $SAVA ADAS-Cog scores over time are all over the place with no clear trend based on longer dosing.
Standard deviations are related directly to Gaussian probability distributions. The stated certain large standard deviations that connote large variations in their data.
Market Data 
Markets 
