In the comments, here's a response that looks like it could have come from members of ecomodder.com:
Quote:
makenakai 
I looked at your test results.
There's only one baseline measurement, so it's really not possible to statistically know the true baseline values, but the ELVIS tests have an average CO% of 0.86 and a standard deviation of 0.25%. This excludes the two measurements with 0.00 CO%, as those should almost certainly be taken as outliers. The baseline CO value is 1.7 standard deviations from the average ELVIS CO measurement.
Given the variation in your data, if you test 50 times you would expect 5 results that are at least 1.7 standard deviations from the mean. Typical statistical significance of a difference between the two conditions  with and without ELVIS  would be assigned if the distance is 1.96 standard deviations or more.
Regarding the hydrocarbons, your baseline is only 0.32 standard deviations from the mean of the ELVIS tests. By all measures, the two conditions can be statistically assumed to be identical.
Do you have more baseline data? That would help firm up the interpretation  it could even help your case.
