A Follow-Up to the First Post

I’ve had my real first idea on how to do some representation since the day of the conference.  For clarity’s sake, I’ll describe my thought process as how I’d do my project up to this point:

There are are a few areas I’d like to try and represent:  the first is race in education.  Since SecondLife allows for a lot of customization of person objects, this shouldn’t be too hard to do at all.  All I really need to do is make 6 (roughly) different looking people along racial lines, and then make as many copies as needed for the project.  I then plan on setting up one of my rooms to look like a school classroom.  This, again, shouldn’t be too hard, because all I would need to do is rez one desk object and then copy it, or even better, just buy one for a paltry $L50 or so (I might have to buy multiple copies, but even 20 desks is just $L1000, or roughly $5 US).  Anything else to set up the classroom would be completely superfluous (like a blackboard, flag, poster, etc) but not hard to add in.  The idea for the actual visualization is this:  pick a state (somehow), and then the people in classroom will rearrange themselves to the proportion of the actual data (in the classroom).  Along with the people, numerical statistics will also be displayed to reinforce the data.  Finally, I came to the conclusion that the best number of people for the classroom is probably 20.  This comes about both empirically and mathematically:  first, most people are used to class sizes of about 20, so this would feel more natural to them.  Second, while 10 makes for a pretty easy breakup (60% white = 6 white people out of 10 total), it’s not really enough to show the distinction between the percentages.  Obviously 25% couldn’t be 2.5 people, but 5 out of 20 is a lot easier to represent.  Thus you could round accordingly:  17.5%-22.49% = 20% or 4 people, 22.5-27.49% = 25% or 5 people, 27.5-32.49% = 30% or 6 people, and so on.

My others ideas are not as concrete.  The two areas I was thinking of so far:  obesity rates in adults, and percentage income.   For now, I cannot decide which would make for a better 3-d visualization in SecondLife.  Though I think obesity rates might be more interesting in data, I think they would be harder to display since the differences in weight aren’t as dramatic as the racial in SecondLife (according to Greta, and my own personal experience).  Also, because I already plan on doing racial data, using people again might feel too similar to that visualization, and thus not be a very big change.  Income data, while not as interesting, is probably easier to model becuase I could use buldings, money, etc. to represent the data, and it would probably feel (and hopefully test) a bit differently.  Otherwise, the idea would be the same as the racial:  pick a state, visualize accordingly, and tack on numbers.  The biggest challenge, for both visualizations, will be to write a script to move around my objects according to the data.

At this point, I was wondering how to display the states.  Though I could use some sort of DropDown menu, this felt rather bland (and clunky from what I’ve used).  During a meeting, Dr. Polack suggested a clickable map.  After pondering this for a while, I’ve come to the conclusion that this is a pretty great idea; it’s very intuitive and easy to follow.  I’ve taken a few classes now, and have learned a bit about building prims and using scripts.  I happened to talk with this amazing self-taught scripter (that’s another story…), and I think I’ve figured out the basic idea on how to make this map.  First, I need to rez a large billboard (which is not hard at all).  Then, I need a to get a map of the US, turn it into a texture, and upload it in to SL.  After applying the texture, I need to create 50 different small “pose ball” type objects to go over each state on the map.  Obviously, the idea is that the user will click the ball for the state of choice.  Each ball will have a very simple script:  send the state ID to another function which will then load and represent the correct data.  This function will be the exact same for every state, so it can be some other hidden object in the room.  From what I’ve learned of object communication so far, this should be pretty easy, because objects can talk and listen over non-chat channels, so the hidden object will be listening for a state, and once a state ball is clicked, it will send the appropriate data (either name “Virginia”, the number 42, or something else) to the displayer function.  This should not actually be that hard to make, and can be in each room.  The end result should be pretty neat looking I think.

That’s basically how I think the visualization aspect of my project is going to unfold.  Unless I find that something is just too hard to do (or maybe even better!), I’ll probably follow along this track for the majority of the time.  Here’s what I need to do to make this project happen:

*****Because I’m still relatively unfamiliar at SecondLife (in the scheme of things), take as many classes as possible, especially ones on scripting and building
*Learn how to texture my US map
*Learn how to upload textures into SecondLife
*Learn how to import Excel data into SecondLife
*Learn how to send information between objects
*Learn how to buy/customize person objects
*Find an appropriate desk/decorations to buy for the classroom
*Figure out a good way to build a room in the first place

That’s really all I can think of for now.  The next couple of days should be pretty uneventful, as I will most likely just continue to keep taking online classes and read more chapters of Tufte’s visualization book(s).   I already feel like I understand SecondLife 100x more than I did 5 days ago, so hopefully I’ll really start to get a good feel in the upcoming week.