An Intelligent Tutoring System (ITS) is an artificially intelligent educational software application that teaches a user skills by giving personalized feedback as the user completes tasks within a problem domain. Despite their popularity, authoring these systems is a labor-intensive process, requiring many different skill sets. A major component of an ITS is the cognitive model. Historically its implementation has required not only cognitive science knowledge, but also programming knowledge as well. To address this challenge, the Extensible Problem Specific Tutor (xPST) was developed for easy authoring of ITSs for existing software and websites. This work develops an xPST authoring tool to simplify the process of xPST authoring by the end user and to help conduct research experiments. It also evaluates the xPST system in terms of the time taken by the users to author successful models. This work also extends xPST framework to enable the creation of generalized tutors in addition to problem specific tutors. To help non-technical military trainers create xPST tutors in game scenarios, this work develops a Torque xPST Driver plugin to enable xPST authoring in Torque 3D game and evaluates authoring in spatial environment scenarios like 3D games using the authoring tool. Finally, this work compares xPST and Cognitive Tutor SDK (another authoring framework) using a fraction addition study and shows that the ratio of training development time to training experience time using xPST is approximately 50% less that that of using Cognitive Tutor SDK. This thesis also shows that there is no significant difference between the “beginner programmer” and “experienced programmer” groups in terms of the time taken to author the tasks using xPST.