Posted by so on June 11, 2005, at 17:39:39
In reply to Re: parties, posted by Dr. Bob on June 11, 2005, at 12:45:04
> And boards might be like that, too. Sometimes better when bigger, but not always...
>
> BobSince "sessions" data from Jan. 05 until now is not gathered using the same methods as "visits" data from before Jan. 05, and...
...since we know there were fewer "sessions" during the past two weeks than during any period since the new protocol was implemented in late Jan. 05, averaging only 86% as many "sessions" this week (3,626) as the weekly average since Jan. 05 (5,596)....
... can you tell us with any certainty whether current "session" averages are more or less than "visits" averages during the four years previous to Jan. 05?
While changed data-collection protocols make accurate measurement difficult, examination of trends before and after the change of protocol suggests a smaller "party" now than before. The monthly average of daily "visits" decreased by an average of 2.66% a month from Dec. 03 through Dec. 04, and since that time, the monthly average of daily "sessions" decreased by an average of 5.58% during the three most recent complete month-long periods under the new data-collection protocol.
The average monthly decrease in "sessions" or "visits" seems to have accelerated in recent months, suggesting the party is continuing to get smaller. Even allowing for uncertainty introduced by the Jan. 05 change in data-collection protocols, only a dramatic increase in page "visits" or "sessions" during the month of Jan. 05 (such as the 38% increase from Jul. 04 - Aug. 04) would tend to support a conclusion that the "party" is now larger, on average, than it was last year. The reported 7.6% decrease from Dec. 04 to Jan. 05 (based on uncertain old-protocol data during the change-over month of Jan.) might be more consistent with the recent decreasing trend than with Dec. to Jan. increases of 14.8%, 17.1% and 3.7% at the beginnings of 2002, 2003 and 2004.
Based on the part of the data we can assume to be internally consistent, do you now think the 60% smaller "party" of Dec. 04 (avg. 7,557 daily visits) was "better" than the largest recorded "party" of Oct. 03 (avg. 18,752 daily visits)? If the "party" now is even smaller, do you think that is even better?
When the "party" was at it's apparent peak in Oct. 03, did you think at that time the larger group presented a better or worse gathering than during previous months? Did you, at that time, consider the increasing number of "visits" evidence of success toward your mission at this forum?
Sources:
http://www.dr-bob.org/stats/2000-05.html
http://www.dr-bob.org:9876/report.cgi?profile=dr-bob.org
poster:so
thread:500533
URL: http://www.dr-bob.org/babble/admin/20050530/msgs/511127.html