Editing Cognitive tutor (section)

== Limitations ==
At this time it is unclear whether Cognitive Tutor is effective at improving student performance.<ref name="US Dept of Ed, 2013"/>  Cognitive Tutor has had some commercial success however, there may be limitations inherently linked to its design and the nature of intelligent tutoring systems.  The following section discusses limitations of Cognitive Tutor which may also apply to other intelligent tutoring systems.

===Curriculum===

At this time, creating a Cognitive Tutor for all subject areas is not practical or economical.  Cognitive Tutor has been used successfully but is still limited to tutoring algebra, computer programming and geometry because these subject areas have an optimal balance of production rules, complexity and maximum benefit to the learner.<ref name="Anderson, J. R., Corbett, A. T., Koedinger, K. R., & Pelletier, R. (1995)">{{cite journal|author=Anderson, J. R.|author2=Corbett, A. T.|author3=Koedinger, K. R.|author4=Pelletier, R.|name-list-style=amp|year=1995|title=Cognitive tutor: Lesson learned|journal=The Journal of the Learning Sciences|volume=4|issue=2|pages=167–207|url=http://www.learnlab.org/opportunities/summer/readings/AndersonEtAlJrnlLearningSci95.pdf|doi=10.1207/s15327809jls0402_2}}</ref><ref name="Scandura, 2012">{{cite journal|author=Scandura, J. M.|year=2012|title=The role of automation in instruction: Recent advances in authorIT and TutorIT solve fundamental problems in developing intelligent tutoring systems|journal=Technology, Instruction, Cognition and Learning|volume=9|pages=3–8|url=http://tutorit.org/Articles/194-TheRoleofAutomationinInstruction-p00-00-Sc.pdf}}</ref>

The focus of Cognitive Tutor development has been the design of the software to teach specific production rules and not on the development of curricular content.  Despite many years of trials, improvements, and a potential to advance learning objectives, the creators continue to rely primarily on outside sources for curricular direction.<ref name="Anderson, J. R., Corbett, A. T., Koedinger, K. R., & Pelletier, R. (1995)"/>

===Design===

The complexity of Cognitive Tutor software requires designers to spend hundreds of hours per instructional hour to create the program.  Despite the time invested, the challenges associated with meeting the needs of the learner within the constraints of the design often result in compromises in flexibility and cognitive fidelity.<ref name="Aleven, 2010">{{cite book|author=Aleven, V.|year=2010|chapter=Rule-Based Cognitive Modeling for Intelligent Tutoring Systems|editor=R. Nkambou|display-editors=etal|title=Advances in Intelligent Tutoring Systems|publisher=SCI 308|pages=33–62}}</ref>

Practicality dictates that designers must choose from a discrete set of methods to teach and support learners.  Limited choices of methods, prompts and hints may be effective in supporting some learners but may conflict with the methods already in use by others.<ref name="Scandura, 2012"/> In addition, it is possible that learners will use the system of prompts and hints to access the answers prematurely thereby advancing through the exercises which may result in them not meeting the learning objectives.

===Model===

The [[cognitive model]], which inspired Cognitive Tutor is based on assumptions about how learning occurs which dictates the chosen instructional methods such as hints, directions and timing of the tutoring prompts.  Given these assumptions, Cognitive Tutor do not account for all the diverse ways human tutors support student learning.<ref name="VanLehn, 2011">{{cite journal|author=VanLehn, K.|year=2011|title=The relative effectiveness of human tutoring, intelligent tutuoring systems and other tutoring systems|journal=Educational Psychologist|volume=46|issue=4|pages=197–221|url=http://www.public.asu.edu/~kvanlehn/Stringent/PDF/EffectivenessOfTutoring_Vanlehn.pdf|doi=10.1080/00461520.2011.611369|s2cid=16188384}}</ref>