New Tech Evaluation and Intelligent Tutoring Systems

Such small things can change the course of people’s lives so dramatically.

..The little boy who says he’s going to such and such high school because that’s where the smart people go…

…The aunt who knew a faculty member of a prestigious school who then arranged a meeting…

…The reception where the student didn’t go talk to the professor after an interview and didn’t get hired as a result…

Stephen Gilbert gave our luncheon lecture today and spoke of his academic career, projects, and networking.

Most recently, Stephen attended a multi-disciplinary conference to discuss visual motion standards. One of the main drawbacks of VR is it’s tendency to cause sickness. Conference attendees—including companies like Google, Sony, Apple, Oculus, and NVIDIA—are considering a labeling system much like ski slopes. How fun would that be?

Apps that are most likely to cause sickness would be labeled as black diamond, and apps that are least likely to cause sickness would be blue. This way, individuals could judge for themselves whether they want to subject themselves to the VR experience.

Would people start to take on the black diamonds as a competition?

Will increased exposure to VR have an effect on susceptibility to motion sickness? In other words, would people find their sea legs?

As for projects, Stephen has dipped his fingers in many areas of research. One of his specialties is the evaluation of emerging technology. In a particular study for Boeing, he asked “Does AR Help with Assembly?”

Three versions of an assembly instruction app were tested—one on PC, one on an iPad, and another on an iPad with AR.

This is quite similar to our CooL:SLiCE project, in that they were testing the existing interface against two new interfaces, one of which included AR.

Participants were asked to assemble a simplified, physical wing. The AR app had the best results for all three measures of time, number of errors (specifically uncorrected errors), and Mean Net Promoter Score. By tracking participants’ movements and what direction they were facing, the study was also able to measure efficiency. The simple mapping and simulation of this type of tracking data is always fun to watch.

The study’s results led to a second study to determine what AR directions are best in what situations. Comparing tunnel and occlusion methods, the study found that results varied by task. Some discussion in our group during the lecture was spurred by questions of how the visualization of each task differed aside from the direction method, and how that might have affected the results.

For example, could a simple animation of a secondary piece moving into place reverse the results?

There is an entire combine simulation at VRAC—seat, screens, wheel and all. In one of Stephen’s studies, he evaluated two apps that wheat and corn farmers would potentially use while driving combines. After modifying the combine to simulate varying degrees of moisture in fields (no easy feat!), they used Empatica E4s and electrodermal activity sensors to gather physiological and biofeedback data.

The data showed an increase in trust over time of one of the apps. The survey results of the study also revealed that the app was more relevant to corn farmers, presumably because moisture is not as much of an issue for wheat.

How do you make the jump from observing people to knowing what people are thinking? This is a core question to Stephen’s multiple intelligent tutoring systems projects.

There have been many ways that attempt to understand what people are thinking, such as sentiment analytics, predictive analytics, and standardized tests (by the way, for an interesting history on testing, the eugenics movement and the term “Asian-American,” consider The Big Test).

Stephen showed several examples of tutoring systems that he has created. There was one for, one for natural language responses (which I thought was the most interesting of all), and more recently one for teams for the military.

The future holds possibilities for games, medical apps, and who knows what else?

In conclusion, Stephen gave anecdotes mixed with charts of how he got into his field and grew his network.

If you apply to grad school, make sure to stop and talk to the professors after your interview! This lesson was learned by a less fortunate applicant.

More good advice in the form of books:

Networking on the Network
How to Succeed in Graduate School
Getting What You Came For

Crimson, gold, and corn are making a comeback! Soon Iowa will witness this new trend, starting somewhere around the VRAC area on ISU campus.

Creating this took long enough… but it’s finally sent to print!



New Tech Evaluation and Intelligent Tutoring Systems — 3 Comments

Leave a Reply