Making do with what you got/ LiPo battery power for voltage, current references

[View[+]Finder]

Attached is a plot of differences in measured voltage readings from an LTZ1000 based reference. The readings were differenced by a Lag+1 such that the ratio was:
Reading(+1)/Reading
then subtract 1 (to remove 100% for clarity as TIN suggested)
and multiply by 100 (for 'PerCent' -not needed) or
multiply by 10^6 and finally
show as PPM

Summary:
mean = 0
StDev = .25
intercept = 0
slope = 0
R-Squared = 0
My conclusion: no serial/auto correlation in the data. Data measured at Lag+10 and Lag+100 show similar values.