GCP Note RB990111
New Year's Analyses

Date: Mon, 11 Jan 1999 18:47:12 -0500
From: Richard Broughton 
Subject: yanya

Roger,

In the parlance of *nix users, "yet another new year's analysis".

OK, First of all, I am a bit new to the field REG analysis, so I may
have made some basic mistakes, so please look this stuff over and, by
all means, let me know if I have screwed up somewhere.

On last Thursday I presented our PentaREG New Year's data to our group.
Having prepared that graph of the Europe and EST New Years, I got to
thinking of the analysis that I would really like to do on the EGG data.
It amounts to taking a logical extension of the EGG analogy, namely,
looking for the equivalent of an averaged evoked potential. (I have
collected PentaREG data on the last dozen or so Seinfeld episodes for
much the same reason, but I have not gotten around to analyzing that.) 
The EGG New Years data gives us that opportunity.

What I thought might be an interesting comparison would be to compute a
composite New Years Midnight plot for those time zones in the world
where the Western New Years is celebrated with gusto, and compare that
with timezones in which there is little celebration (such as those over
the oceans). So, I looked at the time zones and defined a group of 8
zones as "Maximum Celebrating" and another group of 8 zones as "Minimum
Celebrating".

Maximum are: GMT, MET, EET, EAST (Eastern Aus.), Brazil & Argentina,
EST, CST, PST
Minimum are: WAT (mostly ocean), AT (mostly ocean), YST (Yukon and
ocean), HST (Hawaii), NT (Nome and mostly ocean), Dateline west (mostly
ocean), central Russia 4, Dateline East (mostly ocean).
We could probably be a bit more scientific about defining these groups,
but for a first-pass try at this sort of analysis, these seemed
reasonable.

So, I downloaded the EGG data for 12/31/1998 and 1/1/1999 and did the
following:
1. Extracted the 10 minutes of data surrounding NewYear's midnight for
each of the time zones of the globe.
2. Calculated the (Z*Z)-1 for each EGG for every second. Summed those 9
for a composite EGG datum.
3. Summed the values from 2 across the corresponding seconds around
midnight for the 8 maximum and again for the 8 minimum celebrating time
zones.

Thus I ended up with a composite "New Year's Evoked Response" for those
parts of the world in which we could assume there were large numbers of
folks awaiting the stroke of midnight, and a similar ER for those parts
of the world were there were far fewer consciousnesses concerned with
midnight.

The results are plotted in the attached PDF file.  They are interesting.
I am not sure if they are significant, but I have all the data and can
do some further tests or pass it on to someone else on the team. 

Let me know what you think.

Richard