Altering the sampling period shows above to have a must more drastic effect than sampling duration (with six minute data at least). Increasing the time period not only yields more accurate results (as would be expected) but is also much more feasible than trying to find data that is taken at intervals less than six minutes.