?

Log in

No account? Create an account
Researcher Data - brad's life — LiveJournal [entries|archive|friends|userinfo]
Brad Fitzpatrick

[ website | bradfitz.com ]
[ userinfo | livejournal userinfo ]
[ archive | journal archive ]

Researcher Data [Oct. 10th, 2003|12:21 am]
Brad Fitzpatrick
I used to get all excited when researchers came to me for data. They'd want anonymized LJ edge data, or mood data, or stats, or whatever.... social networks, machine learning, etc, etc, etc....

But now it all blurs together and I just feel guilty because I'll never see anything from them and I've stopped being motivated at getting them data.

Sad.

I wish I had a stats department at LJ with a ton of math and statistics geeks doing nothing but cool analysis and putting out geeky press releases and making Zeitgeist-style pages. "LJ Research".
LinkReply

Comments:
[User Picture]From: dakus
2003-10-10 12:28 am (UTC)
Right, I'd guess LJ could put out better numbers than soundscan...

THAT would be interesting...

Wonder what kinda bank SS brings down????
(Reply) (Thread)
[User Picture]From: jwz
2003-10-10 12:49 am (UTC)
This is the part where I think, "surely you could get someone to do that for free!" but of course since people tell me that all the time ("surely you can get someone to be your sysadmin monkey for free drinks!") I know it's not true. How sad.
(Reply) (Thread)
[User Picture]From: jwz
2003-10-10 12:51 am (UTC)

never?

I remember a long time ago you said that you never get to see any results from researchers; have you still never gotten anything?
(Reply) (Thread)
[User Picture]From: brad
2003-10-10 02:00 am (UTC)

Re: never?

There have been a few good nuggets in, but definitely less than I put out.
(Reply) (Parent) (Thread)
[User Picture]From: lisa
2003-10-10 12:57 am (UTC)

intel research labs

what about this one?
(Reply) (Thread)
[User Picture]From: brad
2003-10-10 02:01 am (UTC)

Re: intel research labs

Ah, I'd never seen the results.

I think half the problem is nobody follows up with me, showing me the results.
(Reply) (Parent) (Thread)
[User Picture]From: endquote
2003-10-10 01:04 am (UTC)
Why not just make that data available as some kind of weekly-updated feed to anyone who wants it? Surely someone will make some pretty charts and draw some conclusions from it. Plus by just having it out there, people wouldn't have to bug you about it.

(I know, in all your spare time.)
(Reply) (Thread)
[User Picture]From: brad
2003-10-10 02:02 am (UTC)
Sounds nice in theory, but it's the whole "give an inch, take a foot thing". Once I started releasing data, people would just want more and more. Plus, I'm not sure I trust all people's use of the data. I'm be scared to release something that seems innocent but could somehow be privacy-invading.
(Reply) (Parent) (Thread)
[User Picture]From: bostonsteamer
2003-10-10 09:59 am (UTC)
Aggregate usage/statistical data via RSS or web services would rule!

With a short RFC, I bet any potential privacy concerns could be discovered and resolved quickly.
(Reply) (Parent) (Thread)
[User Picture]From: taral
2003-10-11 03:49 pm (UTC)
Well, you would have gotten data out of me, but you beat me to my own data. :)
(Reply) (Thread)
From: evan
2003-10-12 01:39 pm (UTC)
(you should put this in your memories or userinfo so you can point researchers at it in the future)
(Reply) (Thread)
[User Picture]From: ddelapp
2003-10-12 02:50 pm (UTC)
This quarter I'm taking an Information Design class. Is there any particular information you'd like to see modeled? Perhaps we could work something out.

Check out Understanding USA for an idea of what I mean. Interestingly enough, I just met the author, Richard Saul Wurman this past week and had a chance to speak a bit with him concerning this very topic.
(Reply) (Thread)