<html> <head> <style></style></head> <body class='hmmessage'><div dir='ltr'>THIS MESSAGE IS FOR THE PERSONAL USE OF THE READERS OF THIS LISTSERV AND IS NOT TO BE REPRODUCED FOR ANY OTHER PURPOSE INCLUDING PUBLICATION IN THE SPUT.<BR> <BR>Thank you to Jake Young for setting up the geysertimes website and to Will Boekel and the other people who have helped make the site accessible and useful.<BR> <BR>The other day I overheard someone telling a visitor about the site.  The person then added a comment about usefulness of the average interval computation shown on the website.  As with any other computerized model, information output by the program needs to be used with caution.  Two problems arise with the average and median intervals calculated on geysertimes.  One problem is missing data.  The other problem that arises with some geysers is lumping intervals between various types of activity into a single computation.<BR> <BR>1.  Missing data:  If the sampe size is large enough and the proportion of missing data is small enough then the impact of missing data can be reduced.  However, for example, with Great Fountain, the impact of missing data is not reduced.  On 6/11 when I looked at geysertimes, the average interval listed was 2d16h0m and the median interval listed was 12h51m.  From May 28 (when the road opened) through the 0808E time on 6/11 there were 21 closed intervals and 4 double intervals.  The median of the closed intervals was 11h21m (as compared with 12h51m listed on geyserytimes; the mean of the closed intervals was 11h32m and the overall mean was also 11h32m compared with the 2d16h0m listed on geysertimes.)<BR> <BR>2.  Combining intervals between multiple types of activity:  Castle Geyser demonstrates this problem.  When I looked at geysertimes for Castle activity through the eruption on 6/11 at 1457, the mean interval shown was 12h54m and the median was 13h51m.  When I disaggreated the intervals into major to major versus minor to next eruption back through eruptions on May 24, the mean interval for major to major was 14h09m, with a median of 13h56m.  While the median interval shown on geysertimes is quite close to the median for major to major, the mean interval is over an hour different.<BR> <BR>Summary--use the interval computations on geyser times with caution.  Again thank you to everyone who makes this invaluable site available.<BR> <BR>Lynn Stephens<BR> <BR> <BR> </div></body> </html>