You are here

Objectively Defining Scenario Complexity: Towards Automated, Adaptive Scenario-Based Training

Download pdf | Full Screen View

Date Issued:
2014
Abstract/Description:
Effective Scenario-Based Training (SBT) is sequenced in an efficient trajectory from novice to mastery and is well-grounded in pedagogically sound instructional strategies and learning theory. Adaptive, automated SBT attempts to sequence scenarios according to the performance of the student and implement the sequence without human agency. The source of these scenarios may take the form of a matrix constructed by Instructional Systems Designers (ISD), software engineers or trainers. The domain being instructed may contain procedures or concepts that are easily differentiated thus allowing quick and accurate determination of difficulty. In this instance, the sequencing of the SBT is relatively simple. However, in complex, domain-integrated instructional environments accurate and efficient sequencing may be extremely difficult as ISD, software engineers and trainers, without an objective means to calculate a scenario's complexity must rely on subjectivity. In the Military, where time, fiscal and manpower constraints may lead to ineffective, inefficient and, perhaps, negative training SBT is a growing alternative to live training due to the significant cost avoidance demonstrated by such systems as the United States Marine Corps' (USMC) Abrams Main Battle Tank (M1A1) Advanced Gunnery Training System (AGTS). Even as the practice of simulation training grows, leadership such as the Government Accountability Office asserts that little has been done to demonstrate simulator impact on trainee proficiency. The M1A1 AGTS instructional sub system, the Improved Crew Training Program (ICTP), employs an automated matrix intended to increase Tank Commander (TC) and Gunner (GNR) team proficiency. This matrix is intended to guide the team along a trajectory of ever-increasing scenario difficulty. However, as designed, the sequencing of the matrix is based on subjective evaluation of difficulty, not on empirical or objective calculations of complexity. Without effective, automated SBT that adapts to the performance of the trainee, gaps in combat readiness and fiscal responsibility could grow large.In 2010, the author developed an algorithm intended to computationally define scenario complexity (Dunne, Schatz, Fiore, Martin (&) Nicholson, 2010) and conducted a proof of concept study to determine the algorithm's effectiveness (Dunne, Schatz, Fiore, Nicholson (&) Fowlkes, 2010). Based on results of that study, and follow-on analysis, revisions were made to that Scenario Complexity (SC) algorithm. The purpose of this research was to examine the efficacy of the revised SC algorithm to enable Educators and Trainers, ISDs, and software engineers to objectively and computationally define SC. The research process included a period of instruction for Subject Matter Experts (SME) to receive instruction on how to identify the base variables that comprise SC. Using this knowledge SMEs then determined the values of the scenarios base variables. Once calculated, these values were ranked and compared to the ICTP matrix sequence.Results indicate that the SMEs were very consistent in their ratings of the items across scenario base variables. Due to the highly proceduralized process underlying advanced gunnery skills, this high degree of agreement was expected. However, the significant lack of correlation to the matrix sequencing is alarming and while a recent study has shown the AGTS to increase TC and GNR team proficiency (PM TRASYS, 2014a), this research's findings suggests that redesign of the ICTP matrix is in order.
Title: Objectively Defining Scenario Complexity: Towards Automated, Adaptive Scenario-Based Training.
57 views
8 downloads
Name(s): Dunn, Robert, Author
Sivo, Stephen, Committee Chair
Hoffman, Bobby, Committee Member
Hartshorne, Richard, Committee Member
Bowers, Clint, Committee Member
University of Central Florida, Degree Grantor
Type of Resource: text
Date Issued: 2014
Publisher: University of Central Florida
Language(s): English
Abstract/Description: Effective Scenario-Based Training (SBT) is sequenced in an efficient trajectory from novice to mastery and is well-grounded in pedagogically sound instructional strategies and learning theory. Adaptive, automated SBT attempts to sequence scenarios according to the performance of the student and implement the sequence without human agency. The source of these scenarios may take the form of a matrix constructed by Instructional Systems Designers (ISD), software engineers or trainers. The domain being instructed may contain procedures or concepts that are easily differentiated thus allowing quick and accurate determination of difficulty. In this instance, the sequencing of the SBT is relatively simple. However, in complex, domain-integrated instructional environments accurate and efficient sequencing may be extremely difficult as ISD, software engineers and trainers, without an objective means to calculate a scenario's complexity must rely on subjectivity. In the Military, where time, fiscal and manpower constraints may lead to ineffective, inefficient and, perhaps, negative training SBT is a growing alternative to live training due to the significant cost avoidance demonstrated by such systems as the United States Marine Corps' (USMC) Abrams Main Battle Tank (M1A1) Advanced Gunnery Training System (AGTS). Even as the practice of simulation training grows, leadership such as the Government Accountability Office asserts that little has been done to demonstrate simulator impact on trainee proficiency. The M1A1 AGTS instructional sub system, the Improved Crew Training Program (ICTP), employs an automated matrix intended to increase Tank Commander (TC) and Gunner (GNR) team proficiency. This matrix is intended to guide the team along a trajectory of ever-increasing scenario difficulty. However, as designed, the sequencing of the matrix is based on subjective evaluation of difficulty, not on empirical or objective calculations of complexity. Without effective, automated SBT that adapts to the performance of the trainee, gaps in combat readiness and fiscal responsibility could grow large.In 2010, the author developed an algorithm intended to computationally define scenario complexity (Dunne, Schatz, Fiore, Martin (&) Nicholson, 2010) and conducted a proof of concept study to determine the algorithm's effectiveness (Dunne, Schatz, Fiore, Nicholson (&) Fowlkes, 2010). Based on results of that study, and follow-on analysis, revisions were made to that Scenario Complexity (SC) algorithm. The purpose of this research was to examine the efficacy of the revised SC algorithm to enable Educators and Trainers, ISDs, and software engineers to objectively and computationally define SC. The research process included a period of instruction for Subject Matter Experts (SME) to receive instruction on how to identify the base variables that comprise SC. Using this knowledge SMEs then determined the values of the scenarios base variables. Once calculated, these values were ranked and compared to the ICTP matrix sequence.Results indicate that the SMEs were very consistent in their ratings of the items across scenario base variables. Due to the highly proceduralized process underlying advanced gunnery skills, this high degree of agreement was expected. However, the significant lack of correlation to the matrix sequencing is alarming and while a recent study has shown the AGTS to increase TC and GNR team proficiency (PM TRASYS, 2014a), this research's findings suggests that redesign of the ICTP matrix is in order.
Identifier: CFE0005789 (IID), ucf:50062 (fedora)
Note(s): 2014-12-01
Ph.D.
Education and Human Performance, Dean's Office EDUC
Doctoral
This record was generated from author submitted information.
Subject(s): Scenario-Based Training -- Scenario Complexity -- Adaptive Instruction -- Automated Sequencing
Persistent Link to This Record: http://purl.flvc.org/ucf/fd/CFE0005789
Restrictions on Access: public 2015-06-15
Host Institution: UCF

In Collections