Current Search:   Schatz, Sarah (x)

View All Items

  • CSV Spreadsheet
(1 - 3 of 3)
Leveraging Help Requests in POMDP Intelligent Tutors
Title
Leveraging Help Requests in POMDP Intelligent Tutors.
Creator
Folsom-Kovarik, Jeremiah, Sukthankar, Gita, Schatz, Sarah, Gonzalez, Avelino, Shumaker, Randall, Schatz, Sarah, University of Central Florida
Abstract / Description
Intelligent tutoring systems (ITSs) are computer programs that model individual learners and adapt instruction to help each learner differently. One way ITSs differ from human tutors is that few ITSs give learners a way to ask questions. When learners can ask for help, their questions have the potential to improve learning directly and also act as a new source of model data to help the ITS personalize instruction. Inquiry modeling gives ITSs the ability to answer learner questions and refine...
Show moreIntelligent tutoring systems (ITSs) are computer programs that model individual learners and adapt instruction to help each learner differently. One way ITSs differ from human tutors is that few ITSs give learners a way to ask questions. When learners can ask for help, their questions have the potential to improve learning directly and also act as a new source of model data to help the ITS personalize instruction. Inquiry modeling gives ITSs the ability to answer learner questions and refine their learner models with an inexpensive new input channel.In order to support inquiry modeling, an advanced planning formalism is applied to ITS learner modeling. Partially observable Markov decision processes (POMDPs) differ from more widely used ITS architectures because they can plan complex action sequences in uncertain situations with machine learning. Tractability issues have previously precluded POMDP use in ITS models. This dissertation introduces two improvements, priority queues and observation chains, to make POMDPs scale well and encompass the large problem sizes that real-world ITSs must confront. A new ITS was created to support trainees practicing a military task in a virtual environment. The development of the Inquiry Modeling POMDP Adaptive Trainer (IMP) began with multiple formative studies on human and simulated learners that explored inquiry modeling and POMDPs in intelligent tutoring. The studies suggest the new POMDP representations will be effective in ITS domains having certain common characteristics.Finally, a summative study evaluated IMP's ability to train volunteers in specific practice scenarios. IMP users achieved post-training scores averaging up to 4.5 times higher than users who practiced without support and up to twice as high as trainees who used an ablated version of IMP with no inquiry modeling. IMP's implementation and evaluation helped explore questions about how inquiry modeling and POMDP ITSs work, while empirically demonstrating their efficacy.
Show less
Date Issued
2012
Identifier
CFE0004506, ucf:49262
Format
Document (PDF)
PURL
http://purl.flvc.org/ucf/fd/CFE0004506
Investigating the Effectiveness of Using Part-Task or Whole-Task Training Methods to Facilitate Mindful Abstraction in Uncertain Tasks
Title
Investigating the Effectiveness of Using Part-Task or Whole-Task Training Methods to Facilitate Mindful Abstraction in Uncertain Tasks.
Creator
Killilea, John, Martin, Glenn, Sims, Valerie, Hirumi, Atsusi, Schatz, Sarah, University of Central Florida
Abstract / Description
As the global landscape changes and powers rise and fall, the Contested, Degraded, and Operationally Limited (CDO) environment is likely to be the new normal going forward. Uncertainty variables, such as missing, false, or extra information characterize the CDO environment. A key focus of this dissertation is optimizing training for recognizing these uncertainty variables when training time is limited. This was investigated by either exposing participants to multiple uncertainty variables at...
Show moreAs the global landscape changes and powers rise and fall, the Contested, Degraded, and Operationally Limited (CDO) environment is likely to be the new normal going forward. Uncertainty variables, such as missing, false, or extra information characterize the CDO environment. A key focus of this dissertation is optimizing training for recognizing these uncertainty variables when training time is limited. This was investigated by either exposing participants to multiple uncertainty variables at a time with low doses of each (whole-task training), by exposing singular variables at a time with high doses (part-task training) or using no variables throughout training (control). A key motivator behind this research was Cognitive Load Theory, as mindful abstraction can only occur if there are cognitive resources to spare. Dependent variables, such as time to complete, number correct, task workload, and uncertainty variables identified, were collected.The results revealed that on the transfer task, the part-task condition recorded a significantly lower workload score than the whole-task (and control) condition, while the condition's workload scores were consistent across all training and transfer tasks. In contrast, the control and whole-task condition experienced significant increases in workload during the transfer task. Additionally, the part-task condition participants were able to significantly identify more uncertainty variables on the final task than the whole-task condition and control condition. The part-task condition found the transfer task to be the (")easiest(") in terms of workload, and as there is more opportunity for mindful abstraction if there are more cognitive resources available, it can be stated that based on these results, the part-task training schedule facilitated mindful abstraction more than the other two training schedules. As this was a limited, abstracted, and laboratory experiment, future work should apply the same methodology to applied tasks in a controlled environment to gauge further usefulness of this research.
Show less
Date Issued
2018
Identifier
CFE0007339, ucf:52119
Format
Document (PDF)
PURL
http://purl.flvc.org/ucf/fd/CFE0007339
Bleeding Control Using Multiple Amputee Trauma Trainer in Medical Simulation: Comparison of Movement Versus Non-Movement in Training
Title
Bleeding Control Using Multiple Amputee Trauma Trainer in Medical Simulation: Comparison of Movement Versus Non-Movement in Training.
Creator
Allen, Christine, Kincaid, John, Schatz, Sarah, Allred, Kelly, Talbert, Steven, Sotomayor, Teresita, University of Central Florida
Abstract / Description
Army first responders, specifically Combat Medics and Combat Lifesavers, provide medical intervention while in the field. Didactic as well as hands-on training helps to prepare these first responders, and one module they receive involves bleeding control. First responders are taught to use the Combat Application Tourniquet(&)#174; (CAT(&)#174;) to stop bleeding from limbs subjected to severe injury such as amputation, gunshot, or severe lacerations. A training aid like the Multiple Amputee...
Show moreArmy first responders, specifically Combat Medics and Combat Lifesavers, provide medical intervention while in the field. Didactic as well as hands-on training helps to prepare these first responders, and one module they receive involves bleeding control. First responders are taught to use the Combat Application Tourniquet(&)#174; (CAT(&)#174;) to stop bleeding from limbs subjected to severe injury such as amputation, gunshot, or severe lacerations. A training aid like the Multiple Amputee Trauma Trainer(TM) (MATT(TM)) simulator provides tourniquet training using a lifelike bilateral lower limb amputee. In addition, MATT(TM) combines movement and resistance while the first responder applies the tourniquet, mimicking conditions one would see in a real situation. This research describes tourniquet history, appropriate usage, field tourniquet review, surgical tourniquet, CAT(&)#174; bleeding intervention procedures, bleeding physiology and complications, prehospital tourniquet use in recent conflicts, medical simulation fidelity, and a review of the value of animatronic movement during tourniquet simulation-based training. I then evaluate the effectiveness of animatronic movement during tourniquet training using the Advanced MATT in an experiment using Army first responders. The control group experienced no movement while the experimental group experienced movement when applying a tourniquet during the lab-training. Each group then alternately experienced Advanced MATT movement during an immersive scenario along with fog, strobe lights, and battle sounds. It was hypothesized that 1) In the immersive scenario, the experimental groups (i.e., those who were trained on a moving simulator) would have a faster reaction time as compared to those participants who did not receive training on the moving Advanced MATT simulator; 2) In the lab-based training, the experimental groups would have a slower reaction time; 3) In the immersive scenario, the experimental groups would have a faster tourniquet application time when subjected to movement while in the lab-based training, but the experimental groups would also have a slower tourniquet application time when initially subjected to movement in the laboratory-based training; finally, 4, 5, and 6) Participants who completed lab-based tourniquet training on the Advanced MATT simulation with animatronic movement would report higher perceived realism scores than participants who complete the training on a static version of the Advanced MATT and participants who completed a tourniquet training immersive scenario on the Advanced MATT simulation with movement would report higher perceived realism, presence, and self-efficacy scores than participants who complete the training on a static version of the Advanced MATT. The empirical results show a significant overall training effect of the Advanced MATT simulator (with or without movement). For reaction time and tourniquet application time, involving simulator movement was significant over varying scenarios. A small reduction in reaction and tourniquet application time on the battlefield may be extremely beneficial on the battlefield. Participants who received movement generally gave more positive reactions than those who did not received movement, although these results failed to reach statistical significance. Participants who received movement, followed by a scenario without movement rated the subjective ratings the lowest, suggestive of the lack of movement. Furthermore, despite the order movement was received, no large drops in performance occurred in any condition, indicating that negative training was avoided.
Show less
Date Issued
2011
Identifier
CFE0004468, ucf:49316
Format
Document (PDF)
PURL
http://purl.flvc.org/ucf/fd/CFE0004468