We were instructed to a web site to create our "agent" http://www.gizmoz.com/ I visited the site and was, frankly, disappointed (we were warned!) with the technology. I looked at other people's projects and was not impressed. I tried to make my own and got a million errors. So I bailed and went to Second Life, which I was vaguely familiar with, and worked on an avatar there. Part of my dissatisfaction with gizmoz is probably because Second Life is my frame of reference with avatars.
I did have an interest trip down the wormhole of the web in using gizmoz, tho. I wanted to create a mini-news clip about framing teen pregnancy or teenagers so I did a YouTube search for videos about teenagers. The horrifying teen pregnancy commercial from the UK popped up. (Trying to embed this video was my first "forget it" moment with the application. Later, when I read the journal article that described how the kids' (and adults) would say mean things to the CAs/PAs when then couldn't be helpful, I totally empathized). That commercial -- yikes!
Aside: It was part of the Leicester Teenage Pregnancy and Parenthood Partnership's, paid for by NHS Leicester City which appears to be part of the UK's public health system. Absolutely, positively horrifying and just plain wrong. I'm providing a link against my better judgment; I must focus on this weeks assignment and will postpone my rant.
http://www.youtube.com/watch?v=CRGgWBROuWs&feature=related
I'd been to Second Life http://www.secondlife.com because I saw an amazing presentation at a conference about uses of web 2.0 technology in the public health realm. Rachel Kuchar, who works in the AIDs/HIV prevention group, is also a web 2.0 specialist and she reviewed many different approaches, including CDC's "Island" in Second Life. What really got me interested was a project she described to teach kids about how germs/the flu gets spread. Someone would get infected and pass it along when they talked to someone else. The could only get rid of it by going to see a virtual nurse or doctor. I was impressed by this teaching method because it seemed to actually embed the learning into the kinds of things kids would do in Second Life instead of just "slapping on" some content about it (which is typically the case).
This "slapping on" of content is where we are at in the adolescent health community. If we just create a video, web site, whatever, we're "doing the web." This is far, far from realizing the potential of the technology so I'm always on the look out for truly innovative strategies.
I'm really too much of a novice to comment on how good or not good Second Life is, but the features and capabilities for creating your own avatar are really impressive. Here is Gladys Zeminoba the avatar I created. I picked out the avatars name, gender, race, body size and shape, clothing, facial feature, hair style. I also did her make up, using some sort of special add-on I managed to collect somewhere a long the line. I spent A LOT of time on Gladys' hair and makeup. I honestly wanted to make her look good!
This experience points to how this virtual world exploration (using an avatar, interacting with others) is not a ciick and play operation, yet. I'm fairly competent and parts of it were too challenging for me. The amount of time I'd need to invest to really figure it out and feel comfortable enough to take advantage of all Second Life has to offer doesn't seem worth it. A low ROI, as it were.
This gets back to the question of the week: Are pedagogical Agents (avatars, virtual selves) on the Cutting Edge of Teaching or a Big "Not"?
And the underwhelming answer is: It depends.
In both of the journal articles assigned for this week's class research studies attempted to answer this question as well.
Doering, Veletsianos, Yerasimou's "Conversational Agents and Their Longitudinal Affordances on Communication and Interaction" explored 'which features of virtual characters may influence learning by examining students’ multiple interactions with and responses to conversational agents over time, as well as the effects the students believe these agents had on their learning."
Results: Not so good. Interesting findings about what the participants liked about the technology, but lackluster results specific to the task (educational students trying to create on-line portfolios). This highlights what I believe is the central tension with this technology (and maybe all advances?): what we want or need it to do tends to be based on replacing something we already know or do. And the technology can't do it the way we are currently doing it, so it just doesn't stand up to scrutiny/evaluation. Sometimes the technology catches up and we can shift towards or use the technology replacement. But mostly, I believe the goals are wrong. This orientation limits the potential of the technology. Where is the innovation?
In "When sex, drugs, and violence enter the classroom: Conversations between adolescents and a female pedagogical agent" University of Minnesota scholars Veletsianos, Scharber, and Doering tested a pedagogical agent application with middle school students. It was to assist them with a Google Earth activity. What the kids did instead with the application shocked the heck out of the researchers. Seems like some of the kids were more interested in having Grand Theft Auto style interactions with the agent that getting help with their assignment.
On one hand, this shouldn't be all that surprising. Especially when your agent is a female character interacting with adolescent males. The authors even point out that the development of identity and sexuality is a natural part of adolescence. Even the adults (arguably, older adolescents given the participants' demographics) had semi sexual interactions with the female avatar. In the focus groups at least one reported hanging out with his buddies "talking" to the avatar and there was "chuckling" over this topic. Doering also noted in the Conversational Agent research that the participants wanted to "customize" their avatars based on their own attractions. Hmm. Not hard to imagine students creating their own Angelina Jolie or Brad Pitt avatars. And will that encourage or support their learning or engagement?
These findings all made me curious about how how the avatar looked impacted the interaction. Particularly with the younger participants, I wanted to know how they demographically compared to the avatar that was used.
Most readers of the Veletsianos article were probably appalled by the teenagers' behavior and their negative perceptions of teens reinforced. The knee jerk reaction would be to indict the entire use of the technology and/or insist on teacher monitoring. While I'd agree that there needs to be accountability for bad behaviors, on-going monitoring both defeats some of the purpose of the technology and could ultimately restrict how the kid interacts with it so much that again, the purpose is defeated.
The study made me excited to think about how the teen pregnancy prevention community that I work with could probably program/design excellent content/interactions to deal with sexual inquiries. It would be amazing to be able to convert the kind of inappropriately presented sexual questions/banter into a learning moment about respect, relationships, sexual orientation, gender overall. Instead of an agent that responds "I'm too young for that" when the teenager asks "Are you virgin?" I envision a discussion on what it means to be a virgin, why one decided to have sex, when sex is appropriate, why being a virgin matters, and so on. The teen pregnancy prevention community is already working on web site sex education advisors, even texting based applications. Why not reach into this area? Collaborate with education/instruction technology creators?
My conclusion (based solely on this week's reading and some prior research) is that the avatar technology isn't there yet or we haven't found the right applications yet. I wouldn't go as far as labeling it a "big nothing," but I'm not yet impressed. However, this investigation did inspire one application for me professionally.
Thinking about how AWFUL video, audio and web conferences are, I wondered about ways to use avatars to address the lack of interaction and engagement. The programming capabilities of Second Life and those described in the journal articles seemed impressive. Would it be possible, I wondered, to design a very easy to use application that allows participants in on-line trainings to create simple avatars and populate a virtual classroom or meeting space? Each participant would be able to "speak" through some sort of moderation system and more importantly, respond with non-verbal cues to the speakers and/or other participants. As a participant I could "look" interested, confused, maybe even a little bored.
Having this kind of control would engage the listening participants (which I believe is the big downfall of these apps right now -- so BORING! There's nothing for me to do or engage with so I end up multi-tasking and getting distracted) and give the speaker instant feedback. Again, the ROI question would need to be addressed. The amount of time it would take me to get set up, create my avatar and learn how to use (or as a trainer -- set up the entire system!) would need to outweigh or at least be counterbalanced by how much satisfaction I got from using it. That said, the application I propose here addresses a very real user problem/need and that, in an of itself, might be the key. For the applications discussed in the article, a real alternative exists: ask a real person!
Glynis:
ReplyDeleteWow, that was a well-done post. What was most interesting to me was your information on Second Life. Your experience as a novice was funny and interesting. I wasn't aware that you could spend so much time creating a characters, and do things such as makeup. I was also surprised to learn in that section that public health officials are using tools like this to send messages. Fascinating stuff.
The second part I wanted to comment on was your assessment of the Avatar (Jane, was it?). Maybe this makes me ethnocentric, but I had never considered the cultural competency issues/possibilities the the agent. You are totally right, children of a different culture might respond very differently--most likely negatively--to this character in a business suit and pearls. Excellent point, and one I did not consider at all.
I'm right there with Megan, I didn't think of how a diverse classroom makeup would respond to Jane. That definitely makes a difference in my opinion.
ReplyDeleteI also agree with your "underwhelming answer" that it depends on a lot of different things if an avatar would be appropriate to use in the classroom. I think they could be used effectively if there were some sort of accountability piece included in the assignment.
Glynis,
ReplyDeleteThanks again for having me in your chat group tonight, I had a lot of fun.
But holy smokes, this is quite the blog entry. Although you mention a ton of great stuff in here, I'd like to latch on to one of the first things that you discuss; the idea that you were "disappointed" with the Gizmoz software. Now I've never used Second Life, but I've heard at least that the software affords you WAY more power over the avatar you create in terms of appearance, movements, and so on. But, my question is how high is the barrier to entry with Second Life? Based on your post, although you can get off of the ground with some pretty basic computer knowledge, it sounds like you have to know what you are doing to make a truly life like avatar.
I ask because although extremely limiting, the Gizmoz application is extremely easy to use, and I feel as though this is a must for use in school. Although a few students might have the technical know how to make a mirror like avatar in Second Life, the majority of students will have a hard enough time working with the Gizmoz application.
Anyways, just my two cents. I'm going to have to give Second Life a shot when my First Life calms down a bit!
Thanks again!
-Rick Filipkowski