Learning and Teaching Principles to be Used in SciThink

This list presents a somewhat idealized design philosophy which we will use as a checklist in assessing SciThink games as they develop

Collaborative learning, unsurprisingly, is learning in collaboration with others, usually peers. SciThink games will encourage collaboration within teams and competition between teams.

Computer-adaptive teaching usually refers to teaching that adapts to the learner's level of knowledge. Like the latest computer games, SciThink games will adapt to each player's abilities. They will use simpler language and simpler challenges for players who struggle with basic concepts, and progress in complexity and difficulty with higher levels of proficiency.

Constructivism is the theory that people construct knowledge and meaning from their own experiences, which are often at odds with the "correct" views students are taught in school.

Generative learning is based on constructivism in that the learner constructs new knowledge from prior knowledge. Two simple forms of generative learning are filling in missing letters in a word or missing words in a sentence. Written dialog is a more complex and flexible form of generative expression which allows fuller engagement of players' cognitive faculties. A long-term goal of the SciThink project is therefore to elicit written dialogs between players and the animated figures of great scientists. Research on automated tutoring systems based natural language understanding (NLU) technology suggests that it may be possible to automate much of the scientist side of the dialog, and to use human intervention as a fallback when the NLU is unable to respond adequately. Experimentation will determine how much of the dialog can be automated. The purpose of the NLU is not to replace teachers, but to offload the most repetitive aspects of teaching so they can support more students (players) or focus on helping them with higher level understanding.

Interactive learning is learning through interaction, rather than passive exposure. SciThink games will actively involve players in historical discoveries and modern applications. For example, in "Charles Darwin and the Mystery of Mysteries" players will particate in young Darwin's explorations and discussions. For example, in Chile they will discover marine fossils in mountain ranges and experience an earthquake. Then they will have a chance to use this evidence in discussing the age of the earth with Darwin and his contemporaries. Later in the game, after learning the basics of transmutation (evolution) of species, they will solve modern problems in ecology and medicine. Interactive sections of the game will be interspersed with brief "cut scenes," non-interactive documentary-like sections that fill in socio-historical context, scientific background, and narrative development

Embedded assessment is assessment that is an intrinsic part of learning, and an important element in all kinds of games, including card games, board games, video games, and sports, most of which use some kind of scoring system that is part of the play, rather than a separate test, as in formal schooling. SciThink games will be full of mini-games which will track players' theoretical understanding of scientific principles, and their ability to apply them to practical problems.

Naturalistic learning has no commonly agreed upon definition, except to imply natural settings, as opposed to formal school environments. We use the term to encompass the following idealized set of overlapping features, some of which may sometimes be infeasible:

• Learning in natural settings, both physical and social
• Learning through simulated natural settings when real ones are not available
• Learning through narrative
• Learning through play
• Learning by doing
• Inquiry-based learning, learning designed to solve substantative, real problems.
• Learning by harnessing the natural desire to understand the world and our place in it, AKA natural inquiry.
• Learning through apprenticeship, mentorship, and other forms of extended modeling and dialog between expert and novice
• Socially situated learning, where advancement is marked by increasing responsibility to others and respect from them

Understanding by Design (UbD) is an increasingly popular approach to instructional design developed by educators Grant Wiggins and Jay McTighe. UbD works backwards, first identifying goals, then deciding how to measure progress towards those goals, then determining how to best develop the skills and knowledge needed for those assessments. Like generative learning, UbD views understanding as the ability to transfer knowledge, i.e, apply it flexibly to related but different situations. UbD also avoids two common pitfals in education, as attempting to cover too much material, and failing to provide sufficient context for learners to relate new knowledge to their prior understandings.