Hi Joni,

Thanks for your comments and questions!

Yes, professors at both the broadcasting and receiving institutions help mentor students. The success rates tend to be similar at both the broadcasting and receiving institutions. 

We are constantly striving to improve interaction among all the constituents, professors and students. The difficult part is, of course, breaking down the barriers between the institutions. While we have made progress in this direction, there is still a lot of work to do. Much ultimately depends upon the instructors' abilities to make the classrooms at both ends feel like a single larger classroom despite the distance: Inviting the (reluctant) students at a distance into the broadcasting classroom. Having projects composed of teams from both institutions also helps somewhat but is also not a panacea. Hong, Raphael, Michael, and Michael can provide additional insights.

Dr. Falk, thank you for your comments. 

This is Hong, a Co-PI of the project. To answer your first question, Yes, the students who are learning from the remote classroom have either a Co-Teacher sitting with them when the number of students is large or a TA with them when there is only a couple of students. The Co-Teacher not only helps to address the immediate/emergent need of the students, but also gains an opportunity to serve as a peer review of the teaching, and learn/improve teaching for the next round when the co-teacher takes turns to teach the same course.  The student evaluation data (drop rate) shows that the local students and remote students of the primary teacher have no noticeable difference.  
We have a course website and encourage students to collaborate online with the teammate in remote campuses. But the students mostly prefer to team-up with the local students if they have the choices.  To address this issue, we are building a Computer-Supported Collaborative Learning platform based on AI-technology (webbot) currently. 

I hope that my answers make sense to you. As an evaluator, Mike (Jonathan), probably can answer your second question better.

Thanks for your interest to BOT Dr. Philip,

The web bot is a chatbot built on the Microsoft Bot Framework.  It serves as a virtual TA who monitor the student teamwork under teamwork social media platform such as GroupMe, and Slack, etc. It collects data from the teamwork environment. Therefore, the instructors can evaluate the teamwork process and individual performances besides the team project outcomes.   

The details can be found on the linked article under review. 

Hong 

Dr. Worsley, Thanks for your question and interest. 

The purpose of the Bot Captain is to make the performance and contribution of the teamwork transparent to the instructors as well as the students in the same way as those of athletes are transparent to the coaches and spectators. The system can help to mitigate the two extremes, hitchhikers and heroism. It aims to promote peer collaborative learning and make each teammate more accountable. The Bot Captain logs three types of formative learning assessment data, 1) the student efforts & motivation based on presences and absences of meetings, the roles and tasks, and progress report on each weekly online meeting records, 2) social relational data based on the positive/negative words in communications, 3) cognitive assessment data based on the frequencies of conceptual keywords, and stages of the projects (e.g. initiating a topic thread, topic diverging, topic converging, and resolution). The system sends emails to remind students of the deadlines of the tasks, meetings, etc. and collects data such as task decomposition, the leading role and supporting role for each task. Since the team project counts as 40% of the course grades (homework, quizzes and test count 60%), the scores of the team project consist of two parts:  60% for the summative assessment of the project report, presentation, and codes that all teammates shared the same.  The other 40%, however, is evaluated differently for each individual based on the logged data from Bot Captain, and peer evaluations. The data analytics components have not integrated with the Bot Captain at this stage. The team of students will integrate the two components with another third-party database called Learning Record Store built by the ADL (Advanced Distributed Learning ) group of ARL. The data interoperability is based on xAPI (eXperience API), a JSON (JavaScript Object) format.  I am not sure if I have answered your question. More details can be found on the linked article and future publication. As a product, it will have a debut in fall 2019 working more like an Amazon Mechanical Turk, instead of full automation.   

Thank you. This is what I was wondering.

As Dr. Hong has said, we are still pursuing how best to assessment student outcomes. One methodology involves an approach that involves capturing how a student thinks about a complex problem in the form of an annotated concept map or causal influence diagram and then analyzing that representation in comparison with a reference or expert model (see Spector & Koszalka, 2003; see https://www.nsf.gov/awardsearch/showAward?AWD_ID=0335644). Unfortunately the DEEP software is not longer available. Another method is to examine longer-term impact such as how students progress in their studies and what courses and degrees they pursue. The DEEP software is being revised and we also support the longitudinal approach, although as yet we do not have strong indicatioins other than student and faculty reports on surveys which are a weaker meythod.

Thank you Dr. Liu for the info. Dr. Spector and you answered my question.

Dear Dr. Miller,

It is a very nice question. We really walked around all business troubles in a way similar to that our kids share a ride to school by carpool. We just take turns to teach (drive) and co-teach (free ride). 

More specifically, the students in remote campus (e.g. at Adams State University at Colorado, ASU) always register a course from their own school - ASU (e.g. a special topic MA499/CS499 if the ASU does not have the same title).  But their class is delivered from another university (Embry-Riddle at Daytona, ERAU) through teleconference tools when the teacher teaches his own students face-to-face at ERAU. Typically, a co-teacher is sitting at the remote classroom (at ASU Colorado). The role of the co-teacher is to serve as a peer reviewer of the course, and also learn how to teach a new subject or teach better by learning each other. When the semester ends, the teacher of the face-to-face class at ERAU submit the grades of the remote students to the co-teacher at ASU. Then, the co-teacher will assign (adjust based on the school expectation) the grades of ASU for the remote students at ASU. Next term, the teach and co-teach exchange role for another class, and their students also exchange from face-to-face students to online students. 

Not sure, if this is clear. If you have more questions, please let me know. 

Thanks for the reply. Interesting model.  What sorts of challenges did you encounter with recruiting faculty at other institutions to embrace this model?

Another excellent question. In fact, we are currently setting up a workshop from 4-6 June 2019 for the purposes of disseminating our results, as well as for recruiting new faculty and scaling up our efforts for the next stages of collaborations. Although we have a limited number of spots, we encourage interested faculty to contact us for additional information.