Re: tn41

[Tom Droege <droege@fnal.gov>]

Arne Henden says below:

"Better to show that you are picking up all the known
variables that you should."

I agree with this approach 100%.  The world is expecting us to "go
off half cocked".  We will impress the establishment more if we 
are conservative than if we try to dazzle them with a large number
of "discoveries".  Not that I am not sympathetic to making claims of
exploration.  That is more my natural tendency than to be conservative.
All my discussions with the AAVSO people indicate that they expect us 
to make many false claims and to generate lots of bad data.  We will
do better if we beat the AAVSO with the quality of our data.  This 
should not be difficult for us.

Tom Droege



At 10:58 PM 6/2/98 -0700, you wrote:
>Michael keeps expanding this note, which is an excellent
>description of how to approach the data analysis phase of the
>TASS data.  (It can only improve when the full database is used.)
>  A couple of comments (since my mail to Michael bounces on an
>irregular basis):
>  Your '100 candidates' is based on the V<10 sample only?  If so,
>these should all be well-studied variables since the Tycho
>catalog is complete at this level (amongst other sources of
>information).
>  My experience from the FASTT survey is that more than half of
>the variables that pass tests like W/S are false.  Therefore,
>be really careful before advertising a 'new' variable on the
>poster.  Better to show that you are picking up all the known
>variables that you should.
>  I think Glenn's dataset has ~60K stars.  Figuring normal
>statistics, this means there should be 500-1000 variables in
>the database.  Lots more than you have found so far, but
>since Glenn is the only current observer in the southern
>zone, you will not be able to uniquely ascertain periods less
>than one day.
>  I'm surprised at the large delta between my UBVRI catalog
>and Glenn's data.  If Tycho matches, so should my data.
>I'll try to get some additional fields added before the AAS
>meeting to increase the number of usable stars.
>Arne
>
>