View Single Post
May26-04, 07:38 AM
P: 565
Quote Quote by Hurkyl
If 1/20 is within the experimental error of the setup, then yes, the MM result was "NULL". I've found this table, though am looking for a better online reference. (in particular, one that matches up the numerical data with the precise experimentation)

In any case, for each run, there is significant variation in the data; for example, in the first experiment, the range in the observations in the initial 10 runs is 330, a whopping 36% of the average (913), while in the after data the range is 240, 27% of the average (905). There is simply no way that such large error in observations could confirm a 5% difference as being nonnull.
No the claim was the result was 1/20 of the predicted value. The null results imply No shift within experimenal error.