How to design a system to centralize data? -- using ruby

Hi

I work in a psychology lab which used to test subjects one on one. A
subject would be tested on 1 to 20 tests each generating a summary
value or text data file. Now in addition, we are getting thousands of
responses on single or multiple tests from folks on the web in multiple
websites. We can’t possibly copy/paste that volume of data into an
excel file. So we need to automate.

One thought is to have a centralized depository with both data and data
format so folks can analyse the data easily. Plus a way of filtering
the data: for example extract data from subjects matched on some
criteria… So I think that means a database.

Because the high volume data is accumulated on the web and can be sent
by email, that suggests a webapplication. ruby on rails can probably do
something like that.

Is this something a programmer with a little ruby and rails can design,
or should we consult with some folks to figure out what the
format/system should look like, and if so what kind of folks would be
able to advise us? (folks knowing ruby or rails of course)

I work in a psychology lab which used to test subjects one on one. A

If you’re lab is at harvard as your email suggests, I’d see if there
aren’t some CS students that need a project to complete for their
coursework and talk to them…

-p

On 9/26/06, anne001 [email protected] wrote:

format so folks can analyse the data easily. Plus a way of filtering
the data: for example extract data from subjects matched on some
criteria… So I think that means a database.

Yup.

Because the high volume data is accumulated on the web and can be sent
by email, that suggests a webapplication. ruby on rails can probably do
something like that.

Well, you mention “multiple websites”, as if the data is already being
generated and you need to pick it up and do something with it.
Correct?

In that case, you may not need a webapp per se, since (it sounds like)
the webapps are already in place. If they’re already sending you
auto-generated emails, then you’re set: you just need to hack together
a script to periodically scan the relevant inbox, pull out those
messages, parse them for the data, and insert that data into your db.

Then you’ll probably use a 2nd program to mine that data and do
something useful with it.

Is this something a programmer with a little ruby and rails can design,
or should we consult with some folks to figure out what the
format/system should look like, and if so what kind of folks would be
able to advise us? (folks knowing ruby or rails of course)

I like Philip’s suggestion about finding a CS student or two to do it.
Just post a note on the bulletin board in the CS department and you
should get some help that way.

—John

On Wednesday, September 27, 2006, at 10:05 AM, John G. wrote:

One thought is to have a centralized depository with both data and data
Well, you mention “multiple websites”, as if the data is already being
something useful with it.
—John

Students can be a fine solution to problems like this, but as with most
things, you get what you pay for.

_Kevin
www.sciwerks.com

Thank you for your post,

Unfortunately it is almost impossible to communicate with Computer
Science students at Harvard. I have tried in the past, Harvard has no
intra posting ability, only computer science students can belong to
computer science association, etc.

but we get bright student coming through our courses, so we can hook
students in that way.

Your response is very helpful in that it suggests that for now, we
don’t need to change the way we acquire and store the data, and can
develop the interface seperately. This will reduce the complexity of
the project – if it ain’t broke don’t fix it – .

anne