So today is my first that i got my hands on the Inflow software that Veldis gave us. The software basically analyzes and maps relationships between people based on their commonality. My accomplishment today with the software is I am able to export a database filled with people's profile.
To do this:
Export the existing database to be an excel file and save it as ".CSV" format
Go to inflow, File>import>Node data
There you go, the software will show you all the nodes according on how many records you have in the excel.
The second process to mapping is to gather the "data link". At this point, this is something that I am not clear of and need to wait for Veldis to train us.
Thanks for taking this role Jonathan. As we talk about what we're doing I hope other interns and volunteers will want to add their own time and talent. Or, you'll learn to apply this concept to your own network, in Chicago or in other places. We've a lot to learn.
As you figure out how to create the "data link" which would enable us to show the relationships between the different people in the database, we also need to figure our how Valdis exports the visuals to his blog and other work. Is there a "save" feature that stores projects he is working on? Is there an export feature that turns charts to jpg formats that can be posted in blogs or ppt presentations?
As you learn this you'll be able to create jpgs and upload them to this Ning site to show what we are working on.
Jonathan, my name is Katie Anderson, I'm from Dominican University (Graduate School of Library and Information Science). I'm really interested in helping. I haven't yet used the software, but Dan explained his visions for it. I really look forward to learning the full potential and helping to develop a project for it. I really like how you posted what you learned during your first training. I think if we continue to do so, it will help to share and improve what we know.
Here's the progress I made on Monday, March 29, 2010.
Figure 1: Pre-training map of everyone
This illustrates how we were connected before the InFlow training day with Valdis.
Figure 2: Pre-training map of SNA project team
This is another scenario of Figure 1, simply hiding the nodes/people who won’t be working on the specific SNA project at Tutor/Mentor.
Figure 3: Post-training map of everyone
This illustrates the increased connections after the training day with Valdis.
Figure 4: Post-training map of SNA project team
This hides the people from the previous map who won’t be working on the Tutor/Mentor project.
All users are mapped to network 1 in this case.
An example of of using more than one network could be something like Network 1 = partnerships that existed before the conference; Network 2 = partnerships that were a result of contacts made at the conference; Network 3 = attended the same conference but reported no relationship. We would be able to hide/show/overlap these for different scenarios.
Strength was rated 0-4.
0 = never met before
1 = part of the same project but have not met
2 = only interaction was at InFlow training
3 = have had subsequent interaction, or another connection (for example, same school)
4 = have an existing/ongoing relationship outside this project
I guessed at these relationships because we hadn’t done a survey. In reality, a survey would list all the names of people we’re interested in mapping and ask each participant to indicate/rate the strength of their connection to that person using the above scale. For our scenario, this is easy since the question we can ask easily correlates to a numbered scale (0-4).
In extrapolating to something like mapping the conference participants, I think multi-part questions in a survey will get us what we want.
We could list all participants with a series of checkboxes. For example:
__ I have no relationship with this organization
__ I have plans to partner with this organization as a result of conference connections
__ I have an existing relationship with this organization
If yes: ___ relationship was a result of conference connections
___ relationship was not a result of conference connections
How much interaction do you have with this organization:
__ Less than 6 times/year
__ More than 6 times/year
Translating these questions could be something like: Network 1 = partnerships that existed before the conference; Network 2 = partnerships that were a result of contacts made at the conference; Network 3 = attended the same conference but reported no relationship. We would be able to hide/show/overlap these for different scenarios.
Interaction would correspond to the strength ratings of 0-4.
Attached are the two different .csv files I used for pre- and post-training. The differences between the maps with everyone and just SNA project members was a matter of hiding nodes in InFlow.
Ok guys, so I spent several hours at Cabrini Connections tonight, and developed three maps.
The first two maps that I made examined the relationships between people within our project group. Rather than doing before and after the training, I chose two dates, one being today, March 30, 2010, and one being February 1, 2010, which is somewhat more hypothetically the period I first became aware of and willing to work on the project. For both maps I used the following scale:
1=project participant, but have not met other participants
2=have met & interacted on Ning OR attended the training session
3=have met in person to work on the project
4=have developed a regular, ongoing relationship
*obviously, some of the data is guessed at because I was unable to ask questions to attain the data, but it assumes that everyone met at training except me.
Questions that would work to gather data to answer the above questions:
1. Who have you interacted with on the project?
2. What have you attended pertaining to the project?
3. What is your activity level on Ning (also attainable by looking at Ning)?
SNA Group Relationships 03.30.10
SNA Group Relationships as of 02.01.10
Alright, so after completing those two maps, I wanted to look at the actual organizations that we are affiliated with that this project is bringing together. Because I have not met everyone, I limited this to Anne, Jonathan, Dan, me, and Vladis.
The scale I used was:
1=recent light connection between orgs
2=stronger, but still recent
The map I made turned out like:
I'm mildly not satisfied with this map. As I was leaving this evening, I realized that I should have not connected the university's to InFlow, but used Tutor/Mentor to connect them because I think that's a more accurate representation.
Please give me feedback! Anne, I really thought that you did a good job with yours. I think figuring out the scales and then how to ask questions that give us the right data is key. I think it would be good to start experimenting outside of our teeny group.
Good work Katie. I encourage you and Anne and others to look at the original blog article that Valdis wrote prior to the conference in November. When you update your maps, try to show the connections between Valdis, myself and Jean that were in place prior to Feb. 1 when you first made contact with T/MC.
That's a really interesting article. We could have done the same thing with our group where we go step by step to show how the network is forming. That helps to add some way to analyze the results. Our results are limited because I know I was guessing on some of the relationships that existed for both maps, more so for the Feb. 1 map. Regardless, the maps do show the creation of an inter-connected network, they just don't do a good job showing who has facilitated connections. I think that a more step by step approach, coupled with narrative would assist in developing maps that can be more deeply analyzed.
As April is arriving, I think we may want to start considering what we want to accomplish for the Conference in May. The questions that Anne developed seem like they should work to get the information we need from the organizations. I realize that Anne looked at analyzing them into a 0-4 scale, but I think because there is a relative complexity to the level of relationships, we may want to use a slightly broader scale, InFlow allows us to go to 9, but I was thinking maybe like a 0-6 scale. This accommodates for the different rates of interaction to be shown visually as well.
Anne, I like that the questions are simply check boxes, I think that will make for a more effective and standardized set of data to work with.
Dan, do you have a list of conference attendees that we can reach out to. Because the numbers are going to be larger and it would be nice to have some time to play with the data in InFlow, I think we should get started on this.
On another note (I know this is getting long), I know we talked about mapping the students involved in Cabrini Connections. I think this could be a really neat exhibit for the conference and for explaining social networking and how social networks develop. The following is more brainstorming than a concrete idea, please feel free to respond to it!
As I understand, we have 7th-12th graders at Cabrini Connections. While we will be unable to map the progressive network growth for a single class at this point, simply because the data doesn't exist, I think that it would be useful to compare the networks of say the seventh graders and the 12th graders to show the difference in the networks and how CC helped the network around the students to grow. Therefore, we would need to survey those two classes.
I guess the difference is that what I'm envisioning places a single student at the center, then shows the people that have influence on their life. I'm not quite sure how to do this for a whole class, or if 1 student as an example works.
The survey might look something like rank 0-4 the interactions you have had with the following people:
__People outside your neighborhood
etc. (it would be a longer list)
The number that the students place be our scale.
I'm also thinking, and this just popped into my head, that we could so a map of the relationships between the students, to show the interior relationships that they have built.
This would look more like the maps that we have created for ourselves. I would really like to run a survey especially on the newest students, then repeat it once a year as they go through the program. This (obviously a long term project) would enable us to analyze when different types of relationships form, and also see if there is some sort of 'pop' moment, where the kids really embrace the opportunities open to them, or if it's more of a steady growth.
Okay, that was a lot, and more than enough for now. I look forward to hearing what you all think of my ideas!