On 7/3/2012 7:27 PM, Michael Sokolov wrote:
> what all the pieces of paper and files are for and so on. Once you
> have amassed all this raw data, look for patterns, keep the outliers
> as mementos, and start building your model. The modeling is the
> creative part to be sure, but legwork comes first I think.
So if I had actually *read* your post I would have seen that you said
that you already had data. I still think it's helpful to understand how
people use the data (and how others have modeled it), as others have
said. But you could just dive right in, too, depending on how
structured the data already is. I think the less structured it is, the
more background you need in order to interpret the data meaningfully.
